CIS545Project.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[]},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"markdown","metadata":{"id":"2OJW_QtKLbhA"},"source":["# CIS 5450 Final Project: Amazon Book Reviews \n","_Due: December 15th, 11:59 PM EST_ \n","\n","<br>\n","<center><img src = \"https://cdn.dribbble.com/users/112264/screenshots/1505195/media/4a0c791a7792be31e4367fb0b7b0835e.png?compress=1&resize=400x300&vertical=top\" width= \"300\" height= '200' align =\"center\"/></center>\n","<br>\n"]},{"cell_type":"markdown","metadata":{"id":"kMt0EG1pEv5d"},"source":["# Team Members\n","\n","The team members who worked on this project were **Grace Benner, Naomi Maranga, and Kofi Nightingale** and we were advised by TA **Ernest Ng**\n","\n"]},{"cell_type":"markdown","metadata":{"id":"JvWAeDNhL1-b"},"source":["#Project Goals\n","\n","<div class = \"row\">\n","<div class=\"column\">\n","We intend to study correlations between features of book reviews on Amazon including sentiment analysis on reviews of books across genres, total number of reviews per book, effect of price on book popularity, predicting public sentiment on book authors from reviews of their work(e.g public opinion on figures such as Dr. Seuss or J.K. Rowling etc.)\n","\n","Ultimately, we'd like to predict the number of ratings a book has(ratingsCount) using a variety of features on books(some which are given and some which we'd have to feature engineer.)\n","\n","Some models we're currently considering include logistic regression, KNN, LDA/QDA, and random forest since ratingsCount is a quantitative feature hence lending our project into the regression category. \n","\n","We will also be primarily working with pandas dataframes and operations thereof in conjunction with SQL queries. We will also utilize Apache Spark to delve into the merits of parallelization\n","</div><div class=\"column\">\n","<p class=\"d-flex\" align = \"center\">\n","<img src = \"https://miro.medium.com/max/1400/0*NNTTYmmYvGHvca5S\" height= \"200\" align =\"center\"/>\n","<img src = \"https://ces-sarahboucher.weebly.com/uploads/2/4/7/8/24789555/predict_orig.jpg\" height= \"200\" align =\"center\"/>\n","</p>\n","</div>\n","</div>"]},{"cell_type":"markdown","metadata":{"id":"VccGm4vvCjUP"},"source":["# Value Proposition \n"," We find this project to be interesting since we think that ratingsCount as a feature gives us an overall idea of book popularity. The more ratings say a book has, the more popular it is. We are interested in finding which features predict or could predict a \"best-selling\" book\n","\n","We'd also be interested in looking into collaborative filtering and thus tinker a little bit with recommender systems"]},{"cell_type":"markdown","metadata":{"id":"D3MJvR2zDFJ3"},"source":["# Data Sources and Attributions\n","\n","Our source dataset is acquired from the Kaggle Dataset \"Amazon Book Reviews\": https://www.kaggle.com/datasets/mohamedbakhet/amazon-books-reviews\n","\n","This dataset is built from two main sources:\n"," 1. The first source is https://jmcauley.ucsd.edu/data/amazon/ Citation Ups and downs: Modeling the visual evolution of fashion trends with one-class collaborative filtering R. He, J. McAuley WWW, 2016 Image-based recommendations on styles and substitutes J. McAuley, C. Targett, J. Shi, A. van den Hengel SIGIR, 2015 \n"," 2. The second source is Google books API\n"," The data details file in particular is built by using [google books API](https://developers.google.com/books/docs/overview) to get detailed information about books under consideration\n","\n","**This dataset is in the public domain and is expected to be updated monthly.**\n"]},{"cell_type":"markdown","metadata":{"id":"W0hcZWDcqCUL"},"source":["## Part 0: Libraries, Packages, and General Set Up "]},{"cell_type":"code","source":["#imports \n","import pandas as pd\n","import numpy as np\n","import nltk\n","import matplotlib.pyplot as plt\n","import seaborn as sns"],"metadata":{"id":"iYlrCtf_RRAg","executionInfo":{"status":"ok","timestamp":1671163068251,"user_tz":300,"elapsed":190,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["from google.colab import drive\n","drive.mount('/content/drive')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"QdJ5m7hONbT4","executionInfo":{"status":"ok","timestamp":1671163069509,"user_tz":300,"elapsed":857,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"363d87e9-2fae-4720-b45b-d0845b571474"},"execution_count":8,"outputs":[{"output_type":"stream","name":"stdout","text":["Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"]}]},{"cell_type":"markdown","metadata":{"id":"VzZ6JNnKLplQ"},"source":["\n","\n","# You get a book, you get a book, everyone gets a book!\n","\n","<br>\n","<center><img src = \"https://media.makeameme.org/created/you-get-a-5ba267.jpg\" width= \"500\" align =\"center\"/></center>\n","<br>\n","\n","\n","\n","We'll be loading our data into various dataframes and querying them. We'll also be exploring the following attributes of the book reviews data: \n","- Book-specific metadata such as author names\n","- The books' rankings \n","- Number of reviews which correspond to the number of rows in our dataframe. \n"]},{"cell_type":"markdown","metadata":{"id":"TPm0T5ILfBmH"},"source":["#Part 1: Data Cleaning and EDA\n","\n","We will first load our datasets into a Pandas Dataframe. \n","\n","#### **TODO**\n","- Save the Book Details dataframe to a variable named: <code>books_df</code>\n","- Save the Reviews dataframe to a variable named: <code>reviews_df</code>\n","- Inspect the types of each column in the dataframe\n"]},{"cell_type":"code","source":["books_df = pd.read_csv(\"/content/drive/MyDrive/books_data.csv\")"],"metadata":{"id":"wxwX7rhzN0Dm","executionInfo":{"status":"ok","timestamp":1671163072309,"user_tz":300,"elapsed":2802,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":9,"outputs":[]},{"cell_type":"code","source":["books_df.dtypes"],"metadata":{"id":"WtA8crZdAB83","executionInfo":{"status":"ok","timestamp":1671163072310,"user_tz":300,"elapsed":11,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"e2241cfa-77ac-424d-d91a-0a825ea487b1"},"execution_count":10,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Title             object\n","description       object\n","authors           object\n","image             object\n","previewLink       object\n","publisher         object\n","publishedDate     object\n","infoLink          object\n","categories        object\n","ratingsCount     float64\n","dtype: object"]},"metadata":{},"execution_count":10}]},{"cell_type":"code","source":["books_df.head(5)"],"metadata":{"id":"KEjGrORsBo8P","executionInfo":{"status":"ok","timestamp":1671163072471,"user_tz":300,"elapsed":9,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":372},"outputId":"1a3a58f6-6e9b-4803-b241-a1c0f48a845e"},"execution_count":11,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title  \\\n","0                     Its Only Art If Its Well Hung!   \n","1                           Dr. Seuss: American Icon   \n","2              Wonderful Worship in Smaller Churches   \n","3                      Whispers of the Wicked Saints   \n","4  Nation Dance: Religion, Identity and Cultural ...   \n","\n","                                         description              authors  \\\n","0                                                NaN     ['Julie Strain']   \n","1  Philip Nel takes a fascinating look into the k...       ['Philip Nel']   \n","2  This resource includes twelve principles in un...     ['David R. Ray']   \n","3  Julia Thomas finds her life spinning out of co...  ['Veronica Haddon']   \n","4                                                NaN      ['Edward Long']   \n","\n","                                               image  \\\n","0  http://books.google.com/books/content?id=DykPA...   \n","1  http://books.google.com/books/content?id=IjvHQ...   \n","2  http://books.google.com/books/content?id=2tsDA...   \n","3  http://books.google.com/books/content?id=aRSIg...   \n","4                                                NaN   \n","\n","                                         previewLink  publisher publishedDate  \\\n","0  http://books.google.nl/books?id=DykPAAAACAAJ&d...        NaN          1996   \n","1  http://books.google.nl/books?id=IjvHQsCn_pgC&p...  A&C Black    2005-01-01   \n","2  http://books.google.nl/books?id=2tsDAAAACAAJ&d...        NaN          2000   \n","3  http://books.google.nl/books?id=aRSIgJlq6JwC&d...  iUniverse       2005-02   \n","4  http://books.google.nl/books?id=399SPgAACAAJ&d...        NaN    2003-03-01   \n","\n","                                            infoLink  \\\n","0  http://books.google.nl/books?id=DykPAAAACAAJ&d...   \n","1  http://books.google.nl/books?id=IjvHQsCn_pgC&d...   \n","2  http://books.google.nl/books?id=2tsDAAAACAAJ&d...   \n","3  http://books.google.nl/books?id=aRSIgJlq6JwC&d...   \n","4  http://books.google.nl/books?id=399SPgAACAAJ&d...   \n","\n","                      categories  ratingsCount  \n","0    ['Comics & Graphic Novels']           NaN  \n","1  ['Biography & Autobiography']           NaN  \n","2                   ['Religion']           NaN  \n","3                    ['Fiction']           NaN  \n","4                            NaN           NaN  "],"text/html":["\n","  <div id=\"df-5bd04101-84f9-48e2-a973-6829e4de85ca\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>description</th>\n","      <th>authors</th>\n","      <th>image</th>\n","      <th>previewLink</th>\n","      <th>publisher</th>\n","      <th>publishedDate</th>\n","      <th>infoLink</th>\n","      <th>categories</th>\n","      <th>ratingsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>NaN</td>\n","      <td>['Julie Strain']</td>\n","      <td>http://books.google.com/books/content?id=DykPA...</td>\n","      <td>http://books.google.nl/books?id=DykPAAAACAAJ&amp;d...</td>\n","      <td>NaN</td>\n","      <td>1996</td>\n","      <td>http://books.google.nl/books?id=DykPAAAACAAJ&amp;d...</td>\n","      <td>['Comics &amp; Graphic Novels']</td>\n","      <td>NaN</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>Philip Nel takes a fascinating look into the k...</td>\n","      <td>['Philip Nel']</td>\n","      <td>http://books.google.com/books/content?id=IjvHQ...</td>\n","      <td>http://books.google.nl/books?id=IjvHQsCn_pgC&amp;p...</td>\n","      <td>A&amp;C Black</td>\n","      <td>2005-01-01</td>\n","      <td>http://books.google.nl/books?id=IjvHQsCn_pgC&amp;d...</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>NaN</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Wonderful Worship in Smaller Churches</td>\n","      <td>This resource includes twelve principles in un...</td>\n","      <td>['David R. Ray']</td>\n","      <td>http://books.google.com/books/content?id=2tsDA...</td>\n","      <td>http://books.google.nl/books?id=2tsDAAAACAAJ&amp;d...</td>\n","      <td>NaN</td>\n","      <td>2000</td>\n","      <td>http://books.google.nl/books?id=2tsDAAAACAAJ&amp;d...</td>\n","      <td>['Religion']</td>\n","      <td>NaN</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Whispers of the Wicked Saints</td>\n","      <td>Julia Thomas finds her life spinning out of co...</td>\n","      <td>['Veronica Haddon']</td>\n","      <td>http://books.google.com/books/content?id=aRSIg...</td>\n","      <td>http://books.google.nl/books?id=aRSIgJlq6JwC&amp;d...</td>\n","      <td>iUniverse</td>\n","      <td>2005-02</td>\n","      <td>http://books.google.nl/books?id=aRSIgJlq6JwC&amp;d...</td>\n","      <td>['Fiction']</td>\n","      <td>NaN</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Nation Dance: Religion, Identity and Cultural ...</td>\n","      <td>NaN</td>\n","      <td>['Edward Long']</td>\n","      <td>NaN</td>\n","      <td>http://books.google.nl/books?id=399SPgAACAAJ&amp;d...</td>\n","      <td>NaN</td>\n","      <td>2003-03-01</td>\n","      <td>http://books.google.nl/books?id=399SPgAACAAJ&amp;d...</td>\n","      <td>NaN</td>\n","      <td>NaN</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5bd04101-84f9-48e2-a973-6829e4de85ca')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-5bd04101-84f9-48e2-a973-6829e4de85ca button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-5bd04101-84f9-48e2-a973-6829e4de85ca');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":11}]},{"cell_type":"code","source":["reviews_df = pd.read_csv(\"/content/drive/MyDrive/Books_rating.csv\")"],"metadata":{"id":"T6hxFGOoUw7c","executionInfo":{"status":"ok","timestamp":1671163108496,"user_tz":300,"elapsed":36032,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":12,"outputs":[]},{"cell_type":"code","source":["reviews_df.dtypes"],"metadata":{"id":"7qIDXM5IOdcI","executionInfo":{"status":"ok","timestamp":1671163108496,"user_tz":300,"elapsed":28,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"f62c6ceb-3ad7-4e83-ac4a-5a42a382a5eb"},"execution_count":13,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Id                     object\n","Title                  object\n","Price                 float64\n","User_id                object\n","profileName            object\n","review/helpfulness     object\n","review/score          float64\n","review/time             int64\n","review/summary         object\n","review/text            object\n","dtype: object"]},"metadata":{},"execution_count":13}]},{"cell_type":"code","source":["reviews_df.head(5)"],"metadata":{"id":"3W4pC1VjOfNH","executionInfo":{"status":"ok","timestamp":1671163108496,"user_tz":300,"elapsed":22,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":337},"outputId":"0b9c7bce-9bbb-42d9-8501-95ea275c39f1"},"execution_count":14,"outputs":[{"output_type":"execute_result","data":{"text/plain":["           Id                           Title  Price         User_id  \\\n","0  1882931173  Its Only Art If Its Well Hung!    NaN   AVCGYZL8FQQTD   \n","1  0826414346        Dr. Seuss: American Icon    NaN  A30TK6U7DNS82R   \n","2  0826414346        Dr. Seuss: American Icon    NaN  A3UH4UZ4RSVO82   \n","3  0826414346        Dr. Seuss: American Icon    NaN  A2MVUWT453QH61   \n","4  0826414346        Dr. Seuss: American Icon    NaN  A22X4XUPKF66MR   \n","\n","                          profileName review/helpfulness  review/score  \\\n","0               Jim of Oz \"jim-of-oz\"                7/7           4.0   \n","1                       Kevin Killian              10/10           5.0   \n","2                        John Granger              10/11           5.0   \n","3  Roy E. Perry \"amateur philosopher\"                7/7           4.0   \n","4     D. H. Richards \"ninthwavestore\"                3/3           4.0   \n","\n","   review/time                                   review/summary  \\\n","0    940636800           Nice collection of Julie Strain images   \n","1   1095724800                                Really Enjoyed It   \n","2   1078790400  Essential for every personal and Public Library   \n","3   1090713600  Phlip Nel gives silly Seuss a serious treatment   \n","4   1107993600                           Good academic overview   \n","\n","                                         review/text  \n","0  This is only for Julie Strain fans. It's a col...  \n","1  I don't care much for Dr. Seuss but after read...  \n","2  If people become the books they read and if \"t...  \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...  \n","4  Philip Nel - Dr. Seuss: American IconThis is b...  "],"text/html":["\n","  <div id=\"df-1273369c-5502-42f2-95be-f2613b49df4b\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Id</th>\n","      <th>Title</th>\n","      <th>Price</th>\n","      <th>User_id</th>\n","      <th>profileName</th>\n","      <th>review/helpfulness</th>\n","      <th>review/score</th>\n","      <th>review/time</th>\n","      <th>review/summary</th>\n","      <th>review/text</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>1882931173</td>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>NaN</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>Jim of Oz \"jim-of-oz\"</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>940636800</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0826414346</td>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>NaN</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>Kevin Killian</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>1095724800</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>0826414346</td>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>NaN</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>John Granger</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>1078790400</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>0826414346</td>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>NaN</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>Roy E. Perry \"amateur philosopher\"</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>1090713600</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>0826414346</td>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>NaN</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>D. H. Richards \"ninthwavestore\"</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>1107993600</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1273369c-5502-42f2-95be-f2613b49df4b')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-1273369c-5502-42f2-95be-f2613b49df4b button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-1273369c-5502-42f2-95be-f2613b49df4b');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":14}]},{"cell_type":"markdown","metadata":{"id":"6G0XnVCnv9_h"},"source":["##1.1 Cleaning the Data \n","\n","We'd like to account for NaNs, duplicates or columns that don't give much insight into the data during our EDA step. We could either use imputation or drop columns that are irrelevant to our end goal. \n","P.S: **remember to drop the unnecessary columns before filtering out rows with nulls**\n","\n","\n","\n","<br>\n","\n","#### **TODO: 1.1**\n","\n","- Keep only the relevant columns: \n","- Drop rows that have nulls (e.g. NaN) in them\n","- Make any necessary type casts that would make our data analysis easier\n","- Save the output to `cleaned_books_df` and `cleaned_reviews_df`"]},{"cell_type":"markdown","source":["#### **Cleaning Books Data**"],"metadata":{"id":"KX66Q3BnBS75"}},{"cell_type":"markdown","source":["#####*Drop Unnecessary Columns*"],"metadata":{"id":"4JwBzbUAGeKS"}},{"cell_type":"code","source":["cleaned_books_df = books_df.drop(columns = [\"image\", \"previewLink\", \"infoLink\"])"],"metadata":{"id":"3ETXUzVXPeyo","executionInfo":{"status":"ok","timestamp":1671163108496,"user_tz":300,"elapsed":13,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":15,"outputs":[]},{"cell_type":"markdown","source":["#####*Deal with Null Values*"],"metadata":{"id":"iU78SZZTF8Iz"}},{"cell_type":"code","source":["#Ratings count\n","cleaned_books_df[[\"ratingsCount\"]].describe()"],"metadata":{"id":"TKpTOrR3DwwZ","executionInfo":{"status":"ok","timestamp":1671163108679,"user_tz":300,"elapsed":195,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":300},"outputId":"1c34fca2-5e2a-4265-97d5-59e489183645"},"execution_count":16,"outputs":[{"output_type":"execute_result","data":{"text/plain":["       ratingsCount\n","count  49752.000000\n","mean      21.252975\n","std      201.340431\n","min        1.000000\n","25%        1.000000\n","50%        2.000000\n","75%        5.000000\n","max     4895.000000"],"text/html":["\n","  <div id=\"df-898fc10d-6b29-4c26-8a3a-f7be8d2c2d43\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>ratingsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>49752.000000</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>21.252975</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>201.340431</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>2.000000</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>5.000000</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>4895.000000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-898fc10d-6b29-4c26-8a3a-f7be8d2c2d43')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-898fc10d-6b29-4c26-8a3a-f7be8d2c2d43 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-898fc10d-6b29-4c26-8a3a-f7be8d2c2d43');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":16}]},{"cell_type":"markdown","source":["Since ratings count's minimum value is 1, it is safe to assume that the null values in this column relate to a ratings count of 0. We will impute the null values with 0."],"metadata":{"id":"py8HRPYXD3Rt"}},{"cell_type":"code","source":["cleaned_books_df[\"ratingsCount\"] = cleaned_books_df[\"ratingsCount\"].fillna(0.0)"],"metadata":{"id":"-p-g5rlHD2C3","executionInfo":{"status":"ok","timestamp":1671163108679,"user_tz":300,"elapsed":8,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":17,"outputs":[]},{"cell_type":"code","source":["print(len(cleaned_books_df))\n","print(len(cleaned_books_df.dropna()))"],"metadata":{"id":"Stcui0RTE7PG","executionInfo":{"status":"ok","timestamp":1671163108680,"user_tz":300,"elapsed":9,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"bfbdb6a4-ed58-4dd0-fb7c-fc2ea48a4c9a"},"execution_count":18,"outputs":[{"output_type":"stream","name":"stdout","text":["212404\n","120867\n"]}]},{"cell_type":"markdown","source":["This dataset is very large. Even after dropping the null values we have over 100,000 books to analyze. Since we will be using these data to model with, it is best to just drop the null values now."],"metadata":{"id":"e20HUazQGF0E"}},{"cell_type":"code","source":["cleaned_books_df = cleaned_books_df.dropna().reset_index()"],"metadata":{"id":"7fjZk9HUFsQJ","executionInfo":{"status":"ok","timestamp":1671163108867,"user_tz":300,"elapsed":5,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":19,"outputs":[]},{"cell_type":"markdown","source":["#####*Cleaning Date Published*"],"metadata":{"id":"227fB8FnGn4T"}},{"cell_type":"code","source":["cleaned_books_df[[\"publishedDate\"]].head(10)"],"metadata":{"id":"BIUaDG7vPq-f","executionInfo":{"status":"ok","timestamp":1671163108867,"user_tz":300,"elapsed":4,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"a18d189c-b47c-4391-8eaa-33885cffa79c"},"execution_count":20,"outputs":[{"output_type":"execute_result","data":{"text/plain":["  publishedDate\n","0    2005-01-01\n","1       2005-02\n","2          1996\n","3    2009-01-01\n","4          1995\n","5    1994-02-17\n","6    2018-11-06\n","7    2012-12-06\n","8    2018-02-27\n","9    2010-01-28"],"text/html":["\n","  <div id=\"df-4d5a75ce-a6aa-4352-b6b9-30bb42c30ea0\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>publishedDate</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>2005-01-01</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2005-02</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>1996</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>2009-01-01</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>1995</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>1994-02-17</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>2018-11-06</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>2012-12-06</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>2018-02-27</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>2010-01-28</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4d5a75ce-a6aa-4352-b6b9-30bb42c30ea0')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-4d5a75ce-a6aa-4352-b6b9-30bb42c30ea0 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-4d5a75ce-a6aa-4352-b6b9-30bb42c30ea0');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":20}]},{"cell_type":"markdown","source":["Unfortunately some dates are in the form YYYY-MM-DD, some are in the form YYYY-MM, and others YYYY. Let's create a new column `published_year` since imputation is not as appropriate for the missing months and days."],"metadata":{"id":"mMwnOmNrHBeA"}},{"cell_type":"code","source":["cleaned_books_df[\"published_year\"] = cleaned_books_df[\"publishedDate\"].str[:4]"],"metadata":{"id":"wSkjEx1iGb5V","executionInfo":{"status":"ok","timestamp":1671163108867,"user_tz":300,"elapsed":4,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":21,"outputs":[]},{"cell_type":"code","source":["cleaned_books_df = cleaned_books_df.drop(columns = [\"publishedDate\", \"index\"])"],"metadata":{"id":"eYbN02pCXFZF","executionInfo":{"status":"ok","timestamp":1671163108867,"user_tz":300,"elapsed":5,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":22,"outputs":[]},{"cell_type":"markdown","source":["#####*Cleaning Publisher*"],"metadata":{"id":"qzVRZAO41za-"}},{"cell_type":"markdown","source":["We noticed some inconcistancies with the publisher name. For example, both \"Harper Collins\" and \"HarperCollins\" are values. We are going to concatenate the publisher names to address this."],"metadata":{"id":"3eLUWewn12Au"}},{"cell_type":"code","source":["import re\n","cleaned_books_df[\"publisher\"] = cleaned_books_df[\"publisher\"].apply(lambda x: re.sub('[\\W_]+', '', x))"],"metadata":{"id":"aYhWZHxc2Fj6","executionInfo":{"status":"ok","timestamp":1671163109317,"user_tz":300,"elapsed":453,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":23,"outputs":[]},{"cell_type":"markdown","source":["#####*New Variables using Authors and Categories*"],"metadata":{"id":"WK8RiEzfHmAZ"}},{"cell_type":"markdown","source":["The variables `authors` and `categories` are lists. Create new variables, `num_authors` and `num_categories` to correspond to the number of authors and categories each book has. These may be of value when predicting the ratings count."],"metadata":{"id":"-cfycUEbSz8_"}},{"cell_type":"code","source":["def count_list(str):\n","  count = 1\n","  for i in range(len(str)):\n","    if str[i] == \",\":\n","      count +=1\n","  return count\n","\n","cleaned_books_df[\"num_authors\"] = cleaned_books_df[\"authors\"].apply(lambda x: count_list(x))\n","cleaned_books_df[\"num_categories\"] = cleaned_books_df[\"categories\"].apply(lambda x: count_list(x))"],"metadata":{"id":"lIjATUn8FqJa","executionInfo":{"status":"ok","timestamp":1671163109561,"user_tz":300,"elapsed":245,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":24,"outputs":[]},{"cell_type":"markdown","source":["#### **Cleaning Reviews Data**"],"metadata":{"id":"V_LQBCRnBZzo"}},{"cell_type":"markdown","source":["#####*Drop Unnecessary Columns*"],"metadata":{"id":"mX02hNaKaVL-"}},{"cell_type":"code","source":["cleaned_reviews_df = reviews_df.drop(columns = ['Id', 'profileName', 'review/time'])"],"metadata":{"id":"_k4ypwzadNzv","executionInfo":{"status":"ok","timestamp":1671163109734,"user_tz":300,"elapsed":175,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":25,"outputs":[]},{"cell_type":"markdown","source":["#####*Deal with Null Vlaues*"],"metadata":{"id":"mL_ULLdCdwDS"}},{"cell_type":"markdown","source":["What should we do with the variable `Price`?\n","\n","\n","*   Almost 84% of values for `Price` are null.\n","*   This is too high of a percentage to be able to knowledgably impute the nulls.\n","*   Let's drop `Price`\n","\n","\n","\n","\n","\n"],"metadata":{"id":"QcDv1iaKdOig"}},{"cell_type":"code","source":["cleaned_reviews_df[['Price']].isna().sum() / len(cleaned_reviews_df)"],"metadata":{"id":"OC6qPQhxcOos","executionInfo":{"status":"ok","timestamp":1671163109734,"user_tz":300,"elapsed":3,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"5cd1137d-6722-4c81-ebc2-145bcf87bd70"},"execution_count":26,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Price    0.83961\n","dtype: float64"]},"metadata":{},"execution_count":26}]},{"cell_type":"code","source":["cleaned_reviews_df = cleaned_reviews_df.drop(columns = ['Price'])"],"metadata":{"id":"tV6ctJK2Pff9","executionInfo":{"status":"ok","timestamp":1671163109894,"user_tz":300,"elapsed":161,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":27,"outputs":[]},{"cell_type":"markdown","source":["The only other variable with a high proportion of null values is `User_id`. Since `User_id` will only be used in EDA, and will not be a part of merging the datasets or modeling, we will keep these null values.\n","\n","Otherwise, all other variables have a very low proportion of null values (if they even have any at all), so we will go ahead and drop the rows with nulls in all the columns besides `User_id`."],"metadata":{"id":"hi8xw-S8fOpn"}},{"cell_type":"code","source":["cleaned_reviews_df[['User_id']].isna().sum() / len(cleaned_reviews_df)"],"metadata":{"id":"hT4shybqfXjh","executionInfo":{"status":"ok","timestamp":1671163110402,"user_tz":300,"elapsed":347,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"8d9d03bc-b382-4d51-aec3-61e21c57d104"},"execution_count":28,"outputs":[{"output_type":"execute_result","data":{"text/plain":["User_id    0.187262\n","dtype: float64"]},"metadata":{},"execution_count":28}]},{"cell_type":"code","source":["cleaned_reviews_df = cleaned_reviews_df.dropna(subset = ['Title', 'review/helpfulness',\n","                                                         'review/score', 'review/summary', 'review/text'])"],"metadata":{"id":"e5_rWzChgD4I","executionInfo":{"status":"ok","timestamp":1671163111644,"user_tz":300,"elapsed":1244,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":29,"outputs":[]},{"cell_type":"markdown","source":["#####*Rename Columns* "],"metadata":{"id":"8Y9IPYwViN67"}},{"cell_type":"code","source":["cleaned_reviews_df = cleaned_reviews_df.rename(columns = {\"review/helpfulness\":\"review_helpfulness_rating\",\n","                                                          \"review/score\":\"review_score\",\n","                                                          \"review/summary\": \"review_summary\",\n","                                                          \"review/text\":\"review_text\"})"],"metadata":{"id":"KHZaLjWXiWbh","executionInfo":{"status":"ok","timestamp":1671163111644,"user_tz":300,"elapsed":2,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":30,"outputs":[]},{"cell_type":"markdown","source":["#####*Create a Percent Helpful Rating Column* "],"metadata":{"id":"pBqz7vHbiXgo"}},{"cell_type":"markdown","source":["There is a column in the reviews dataframe called `review_helpfulness_rating` (originally `review/helpfulness`). This column gives the number of people who found the review helpful (x) over the number of people who rated this review (y) in the format x/y. \n","\n","We will create a new column `review_percent_helpful` which finds the percentage of helpfulness for the given review. \n","\n","*   If there are zero ratings for the review (i.e.  `review_helpfulness_rating` == '0/0'), set the `review_percent_helpful` to 0"],"metadata":{"id":"l_LJiwlym6JH"}},{"cell_type":"code","source":["def percent_helpful(str):\n","  items = str.split(\"/\")\n","  if (int(items[1]) == 0):\n","    percent_helpful = 0.0\n","  else:\n","    percent_helpful = int(items[0]) / int(items[1]) * 100\n","  return round(percent_helpful, 2)\n","\n","cleaned_reviews_df[\"review_percent_helpful\"] = cleaned_reviews_df[\"review_helpfulness_rating\"].apply(lambda x: percent_helpful(x))"],"metadata":{"id":"Uhhhq8_Rji7J","executionInfo":{"status":"ok","timestamp":1671163115009,"user_tz":300,"elapsed":3196,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":31,"outputs":[]},{"cell_type":"markdown","source":["#####*Create a Categorical Review Score Column* "],"metadata":{"id":"jggONW2apSfK"}},{"cell_type":"markdown","source":["We will classify a positive review score as having a `review_score` greater than or equal to 4.0. \n","\n","\n","*   If `review_score` >= 4.0 then `cat_review_score` = 'positive'\n","*   If 2.0 < `review_score` < 3.0 then `cat_review_score` = 'neutral'\n","*   If `review_score` <= 2.0 then `cat_review_score` = 'negative'\n","\n","\n"],"metadata":{"id":"670pGrKwplK-"}},{"cell_type":"code","source":["def cat_review(review_score):\n","  if review_score >= 4.0:\n","    cat_review = \"positive\"\n","  elif review_score > 2.0:\n","    cat_review = \"neutral\"\n","  else:\n","    cat_review = \"negative\"\n","\n","  return cat_review\n","\n","\n","cleaned_reviews_df[\"positive_review_score\"] = cleaned_reviews_df[\"review_score\"].apply(lambda x: cat_review(x))"],"metadata":{"id":"Ic512CCAo4xN","executionInfo":{"status":"ok","timestamp":1671163115627,"user_tz":300,"elapsed":626,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":32,"outputs":[]},{"cell_type":"markdown","metadata":{"id":"kYp9fW_SvG3g"},"source":["### 1.1.2 Our Sandbox \n","\n","In our little sandbox, we will do some high-level exploratory data analysis (EDA) to familiarize ourselves with the books and reviews datasets, after the files have been cleaned. Some tools we've found useful in the past include: \n","- `df.head()`\n","- `df.describe()`\n","- `df.info()`\n","- `Series.unique()`\n","\n","<br>\n","<center><img src = \"https://res.cloudinary.com/hy4kyit2a/f_auto,fl_lossy,q_70/learn/modules/nonprofit-success-pack-maintenance/set-up-a-sandbox-in-your-salesforce-org/images/a8e6e5b50adddd6a6baa82854889c0fb_salesforcelandian-npsp-scenes-image-2-building-sand-castles-2.png\" width= \"300\" height='300' align =\"center\"/></center>\n","<br>\n"]},{"cell_type":"markdown","source":["#### **Books Data**"],"metadata":{"id":"Vm_d-OlOWMnV"}},{"cell_type":"code","source":["cleaned_books_df.head(5)"],"metadata":{"id":"TxvQbeRgWUHV","executionInfo":{"status":"ok","timestamp":1671163115627,"user_tz":300,"elapsed":11,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":302},"outputId":"8c4a7ac1-a020-4763-aa91-33359125edb1"},"execution_count":33,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title  \\\n","0                           Dr. Seuss: American Icon   \n","1                      Whispers of the Wicked Saints   \n","2  The Church of Christ: A Biblical Ecclesiology ...   \n","3                           Saint Hyacinth of Poland   \n","4  Rising Sons and Daughters: Life Among Japan's ...   \n","\n","                                         description  \\\n","0  Philip Nel takes a fascinating look into the k...   \n","1  Julia Thomas finds her life spinning out of co...   \n","2  In The Church of Christ: A Biblical Ecclesiolo...   \n","3  The story for children 10 and up of St. Hyacin...   \n","4  Wardell recalls his experience as a foreign st...   \n","\n","                    authors              publisher  \\\n","0            ['Philip Nel']                ACBlack   \n","1       ['Veronica Haddon']              iUniverse   \n","2      ['Everett Ferguson']  WmBEerdmansPublishing   \n","3  ['Mary Fabyan Windeatt']            TanBooksPub   \n","4        ['Steven Wardell']      PlymptonPressIntl   \n","\n","                      categories  ratingsCount published_year  num_authors  \\\n","0  ['Biography & Autobiography']           0.0           2005            1   \n","1                    ['Fiction']           0.0           2005            1   \n","2                   ['Religion']           5.0           1996            1   \n","3  ['Biography & Autobiography']           0.0           2009            1   \n","4             ['Social Science']           0.0           1995            1   \n","\n","   num_categories  \n","0               1  \n","1               1  \n","2               1  \n","3               1  \n","4               1  "],"text/html":["\n","  <div id=\"df-5ba9d569-ac8b-4c6c-b2e0-c91f9c60c466\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>description</th>\n","      <th>authors</th>\n","      <th>publisher</th>\n","      <th>categories</th>\n","      <th>ratingsCount</th>\n","      <th>published_year</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>Philip Nel takes a fascinating look into the k...</td>\n","      <td>['Philip Nel']</td>\n","      <td>ACBlack</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>0.0</td>\n","      <td>2005</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Whispers of the Wicked Saints</td>\n","      <td>Julia Thomas finds her life spinning out of co...</td>\n","      <td>['Veronica Haddon']</td>\n","      <td>iUniverse</td>\n","      <td>['Fiction']</td>\n","      <td>0.0</td>\n","      <td>2005</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>The Church of Christ: A Biblical Ecclesiology ...</td>\n","      <td>In The Church of Christ: A Biblical Ecclesiolo...</td>\n","      <td>['Everett Ferguson']</td>\n","      <td>WmBEerdmansPublishing</td>\n","      <td>['Religion']</td>\n","      <td>5.0</td>\n","      <td>1996</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Saint Hyacinth of Poland</td>\n","      <td>The story for children 10 and up of St. Hyacin...</td>\n","      <td>['Mary Fabyan Windeatt']</td>\n","      <td>TanBooksPub</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>0.0</td>\n","      <td>2009</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Rising Sons and Daughters: Life Among Japan's ...</td>\n","      <td>Wardell recalls his experience as a foreign st...</td>\n","      <td>['Steven Wardell']</td>\n","      <td>PlymptonPressIntl</td>\n","      <td>['Social Science']</td>\n","      <td>0.0</td>\n","      <td>1995</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5ba9d569-ac8b-4c6c-b2e0-c91f9c60c466')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-5ba9d569-ac8b-4c6c-b2e0-c91f9c60c466 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-5ba9d569-ac8b-4c6c-b2e0-c91f9c60c466');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":33}]},{"cell_type":"code","source":["cleaned_books_df.info()"],"metadata":{"id":"0k1pLMduWkl0","executionInfo":{"status":"ok","timestamp":1671163115830,"user_tz":300,"elapsed":211,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"ffd35462-d9e5-4051-a3b4-9a2ac9fcecbd"},"execution_count":34,"outputs":[{"output_type":"stream","name":"stdout","text":["<class 'pandas.core.frame.DataFrame'>\n","RangeIndex: 120867 entries, 0 to 120866\n","Data columns (total 9 columns):\n"," #   Column          Non-Null Count   Dtype  \n","---  ------          --------------   -----  \n"," 0   Title           120867 non-null  object \n"," 1   description     120867 non-null  object \n"," 2   authors         120867 non-null  object \n"," 3   publisher       120867 non-null  object \n"," 4   categories      120867 non-null  object \n"," 5   ratingsCount    120867 non-null  float64\n"," 6   published_year  120867 non-null  object \n"," 7   num_authors     120867 non-null  int64  \n"," 8   num_categories  120867 non-null  int64  \n","dtypes: float64(1), int64(2), object(6)\n","memory usage: 8.3+ MB\n"]}]},{"cell_type":"code","source":["cleaned_books_df.describe()"],"metadata":{"id":"OruJd1VmWmMj","executionInfo":{"status":"ok","timestamp":1671163115835,"user_tz":300,"elapsed":20,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":300},"outputId":"a2b6c315-3b99-4939-b7dd-356a91f3191d"},"execution_count":35,"outputs":[{"output_type":"execute_result","data":{"text/plain":["        ratingsCount    num_authors  num_categories\n","count  120867.000000  120867.000000   120867.000000\n","mean        8.105033       1.313568        1.021056\n","std       127.075136       0.958524        0.150384\n","min         0.000000       1.000000        1.000000\n","25%         0.000000       1.000000        1.000000\n","50%         0.000000       1.000000        1.000000\n","75%         1.000000       1.000000        1.000000\n","max      4895.000000     110.000000        5.000000"],"text/html":["\n","  <div id=\"df-98081ace-a7d6-426b-8c6e-cf0424222d45\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>ratingsCount</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>120867.000000</td>\n","      <td>120867.000000</td>\n","      <td>120867.000000</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>8.105033</td>\n","      <td>1.313568</td>\n","      <td>1.021056</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>127.075136</td>\n","      <td>0.958524</td>\n","      <td>0.150384</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>0.000000</td>\n","      <td>1.000000</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>0.000000</td>\n","      <td>1.000000</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>0.000000</td>\n","      <td>1.000000</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>1.000000</td>\n","      <td>1.000000</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>4895.000000</td>\n","      <td>110.000000</td>\n","      <td>5.000000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-98081ace-a7d6-426b-8c6e-cf0424222d45')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-98081ace-a7d6-426b-8c6e-cf0424222d45 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-98081ace-a7d6-426b-8c6e-cf0424222d45');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":35}]},{"cell_type":"markdown","source":["#### **Reviews Data**"],"metadata":{"id":"25SHA6-0WKNG"}},{"cell_type":"code","source":["cleaned_reviews_df.head(5)"],"metadata":{"id":"8xMWH9msWUn8","executionInfo":{"status":"ok","timestamp":1671163115836,"user_tz":300,"elapsed":15,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":302},"outputId":"b719ccf1-5aba-41e7-eb3b-abb2fd63c943"},"execution_count":36,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score  \n","0              positive  \n","1              positive  \n","2              positive  \n","3              positive  \n","4              positive  "],"text/html":["\n","  <div id=\"df-323a101d-b18e-4652-a406-0a4fe8afc803\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-323a101d-b18e-4652-a406-0a4fe8afc803')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-323a101d-b18e-4652-a406-0a4fe8afc803 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-323a101d-b18e-4652-a406-0a4fe8afc803');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":36}]},{"cell_type":"code","source":["cleaned_reviews_df.info()"],"metadata":{"id":"GalpMurRoZMk","executionInfo":{"status":"ok","timestamp":1671163115836,"user_tz":300,"elapsed":13,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"6691fcb2-267b-4ed0-a857-19ae1590ffea"},"execution_count":37,"outputs":[{"output_type":"stream","name":"stdout","text":["<class 'pandas.core.frame.DataFrame'>\n","Int64Index: 2999746 entries, 0 to 2999999\n","Data columns (total 8 columns):\n"," #   Column                     Dtype  \n","---  ------                     -----  \n"," 0   Title                      object \n"," 1   User_id                    object \n"," 2   review_helpfulness_rating  object \n"," 3   review_score               float64\n"," 4   review_summary             object \n"," 5   review_text                object \n"," 6   review_percent_helpful     float64\n"," 7   positive_review_score      object \n","dtypes: float64(2), object(6)\n","memory usage: 206.0+ MB\n"]}]},{"cell_type":"code","source":["cleaned_reviews_df.describe()"],"metadata":{"id":"SFX7JsMUobdA","executionInfo":{"status":"ok","timestamp":1671163116343,"user_tz":300,"elapsed":515,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":300},"outputId":"744e2f6e-adc9-49ea-a3a7-88be6d23e87c"},"execution_count":38,"outputs":[{"output_type":"execute_result","data":{"text/plain":["       review_score  review_percent_helpful\n","count  2.999746e+06            2.999746e+06\n","mean   4.215262e+00            4.924238e+01\n","std    1.203067e+00            4.309176e+01\n","min    1.000000e+00            0.000000e+00\n","25%    4.000000e+00            0.000000e+00\n","50%    5.000000e+00            5.000000e+01\n","75%    5.000000e+00            1.000000e+02\n","max    5.000000e+00            3.000000e+02"],"text/html":["\n","  <div id=\"df-f18e11c2-7c64-4736-81cc-ad1cd8befd62\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>review_score</th>\n","      <th>review_percent_helpful</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>count</th>\n","      <td>2.999746e+06</td>\n","      <td>2.999746e+06</td>\n","    </tr>\n","    <tr>\n","      <th>mean</th>\n","      <td>4.215262e+00</td>\n","      <td>4.924238e+01</td>\n","    </tr>\n","    <tr>\n","      <th>std</th>\n","      <td>1.203067e+00</td>\n","      <td>4.309176e+01</td>\n","    </tr>\n","    <tr>\n","      <th>min</th>\n","      <td>1.000000e+00</td>\n","      <td>0.000000e+00</td>\n","    </tr>\n","    <tr>\n","      <th>25%</th>\n","      <td>4.000000e+00</td>\n","      <td>0.000000e+00</td>\n","    </tr>\n","    <tr>\n","      <th>50%</th>\n","      <td>5.000000e+00</td>\n","      <td>5.000000e+01</td>\n","    </tr>\n","    <tr>\n","      <th>75%</th>\n","      <td>5.000000e+00</td>\n","      <td>1.000000e+02</td>\n","    </tr>\n","    <tr>\n","      <th>max</th>\n","      <td>5.000000e+00</td>\n","      <td>3.000000e+02</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f18e11c2-7c64-4736-81cc-ad1cd8befd62')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-f18e11c2-7c64-4736-81cc-ad1cd8befd62 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-f18e11c2-7c64-4736-81cc-ad1cd8befd62');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":38}]},{"cell_type":"code","source":["len(cleaned_reviews_df['Title'].unique())"],"metadata":{"id":"nJuWWhM4oe2J","executionInfo":{"status":"ok","timestamp":1671163116709,"user_tz":300,"elapsed":387,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"5116c429-154b-4fc6-b25c-f8d593d12969"},"execution_count":39,"outputs":[{"output_type":"execute_result","data":{"text/plain":["212401"]},"metadata":{},"execution_count":39}]},{"cell_type":"code","source":["len(cleaned_reviews_df['User_id'].unique())"],"metadata":{"id":"MDJw9bCVovO1","executionInfo":{"status":"ok","timestamp":1671163117556,"user_tz":300,"elapsed":850,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"f3ba794b-4c9f-42f4-83c3-d1e7fbe5f220"},"execution_count":40,"outputs":[{"output_type":"execute_result","data":{"text/plain":["1008950"]},"metadata":{},"execution_count":40}]},{"cell_type":"markdown","metadata":{"id":"GDWpC6BswHBp"},"source":["##1.2 EDA with Books Data\n","\n","Since we have cleaned our datasets, and examined them on a high-level, it is now time to do some exploratory data analysis (EDA) with our data. Let's first take a look at `cleaned_books_df` and answer some interesting questions about the data, as well as visualize. \n","\n","We eventually will be using the `ratingsCount` variable as the label in some ML models, so this EDA will mostly focus on how other variables relate to `ratingsCount`.\n"]},{"cell_type":"markdown","source":["#####*Top 10 Most Rated Books*"],"metadata":{"id":"dJCZGkjwLlyt"}},{"cell_type":"code","source":["books_df = cleaned_books_df[['Title', 'authors', 'ratingsCount']].sort_values(by='ratingsCount', ascending = False).head(10).reset_index().drop(columns = 'index')\n","books_df['authors'] = books_df['authors'].apply(lambda x: x[2:len(x) -2])\n","books_df"],"metadata":{"id":"DiRzbiGGJ-9g","executionInfo":{"status":"ok","timestamp":1671163117559,"user_tz":300,"elapsed":37,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"b47bcfee-5b18-44cf-e69b-7c1ffa18ede3"},"execution_count":41,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title            authors  \\\n","0                                      The Alchemist       Paulo Coelho   \n","1                                           Unbroken  Laura Hillenbrand   \n","2  A Mission to Millions; The Story of Ernie Alle...  Laura Hillenbrand   \n","3  Blue Like Jazz: Nonreligious Thoughts on Chris...      Donald Miller   \n","4                                            Eclipse    Stephenie Meyer   \n","5                                     Redeeming Love    Francine Rivers   \n","6                                A Walk in the Woods        Bill Bryson   \n","7  The Tipping Point: How Little Things Can Make ...   Malcolm Gladwell   \n","8  Travel Light, Eat Heavy Along the Appalachian ...        Bill Bryson   \n","9                                          Henrietta     Rebecca Skloot   \n","\n","   ratingsCount  \n","0        4895.0  \n","1        4572.0  \n","2        4572.0  \n","3        4562.0  \n","4        4392.0  \n","5        4249.0  \n","6        4234.0  \n","7        4218.0  \n","8        4161.0  \n","9        4067.0  "],"text/html":["\n","  <div id=\"df-84b7792d-beab-48cc-8dea-d93ab73d073c\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>authors</th>\n","      <th>ratingsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>The Alchemist</td>\n","      <td>Paulo Coelho</td>\n","      <td>4895.0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Unbroken</td>\n","      <td>Laura Hillenbrand</td>\n","      <td>4572.0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>A Mission to Millions; The Story of Ernie Alle...</td>\n","      <td>Laura Hillenbrand</td>\n","      <td>4572.0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Blue Like Jazz: Nonreligious Thoughts on Chris...</td>\n","      <td>Donald Miller</td>\n","      <td>4562.0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Eclipse</td>\n","      <td>Stephenie Meyer</td>\n","      <td>4392.0</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>Redeeming Love</td>\n","      <td>Francine Rivers</td>\n","      <td>4249.0</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>A Walk in the Woods</td>\n","      <td>Bill Bryson</td>\n","      <td>4234.0</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>The Tipping Point: How Little Things Can Make ...</td>\n","      <td>Malcolm Gladwell</td>\n","      <td>4218.0</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Travel Light, Eat Heavy Along the Appalachian ...</td>\n","      <td>Bill Bryson</td>\n","      <td>4161.0</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>Henrietta</td>\n","      <td>Rebecca Skloot</td>\n","      <td>4067.0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-84b7792d-beab-48cc-8dea-d93ab73d073c')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-84b7792d-beab-48cc-8dea-d93ab73d073c button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-84b7792d-beab-48cc-8dea-d93ab73d073c');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":41}]},{"cell_type":"markdown","source":["#####*Top 10 Most Rated Publishers*"],"metadata":{"id":"CxzXu3zXMdof"}},{"cell_type":"code","source":["pub_df = cleaned_books_df[['publisher', 'ratingsCount']].groupby(by='publisher').sum().sort_values(by='ratingsCount', ascending = False).head(10).reset_index()\n","pub_df"],"metadata":{"id":"jGN1SJ_MMdD7","executionInfo":{"status":"ok","timestamp":1671163117560,"user_tz":300,"elapsed":26,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"068ac7e7-df04-4b34-d737-bc7a3c74ed2b"},"execution_count":42,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                   publisher  ratingsCount\n","0                    Penguin       88714.0\n","1              HarperCollins       73759.0\n","2           SimonandSchuster       51226.0\n","3                RandomHouse       48706.0\n","4                    Vintage       40755.0\n","5  KnopfBooksforYoungReaders       27173.0\n","6                  Macmillan       23341.0\n","7                LittleBrown       17399.0\n","8                 HachetteUK       15430.0\n","9                     Bantam       14825.0"],"text/html":["\n","  <div id=\"df-ea0c6774-255e-450a-8239-6e628cb4d801\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>publisher</th>\n","      <th>ratingsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Penguin</td>\n","      <td>88714.0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>HarperCollins</td>\n","      <td>73759.0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>SimonandSchuster</td>\n","      <td>51226.0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>RandomHouse</td>\n","      <td>48706.0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Vintage</td>\n","      <td>40755.0</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>KnopfBooksforYoungReaders</td>\n","      <td>27173.0</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Macmillan</td>\n","      <td>23341.0</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>LittleBrown</td>\n","      <td>17399.0</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>HachetteUK</td>\n","      <td>15430.0</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>Bantam</td>\n","      <td>14825.0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ea0c6774-255e-450a-8239-6e628cb4d801')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-ea0c6774-255e-450a-8239-6e628cb4d801 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-ea0c6774-255e-450a-8239-6e628cb4d801');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":42}]},{"cell_type":"markdown","source":["#####*Top 10 Most Rated Categories*"],"metadata":{"id":"H6dZ1uaZNCTz"}},{"cell_type":"code","source":["cats_df = cleaned_books_df[['categories', 'ratingsCount']].groupby(by='categories').sum().sort_values(by='ratingsCount', ascending = False).head(10).reset_index()\n","cats_df['categories'] = cats_df['categories'].apply(lambda x: x[2:len(x) -2])\n","cats_df"],"metadata":{"id":"5biwmMVsNQif","executionInfo":{"status":"ok","timestamp":1671163117561,"user_tz":300,"elapsed":25,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"3f6fc9c2-53d8-4474-d9fb-db1bc307ade7"},"execution_count":43,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                  categories  ratingsCount\n","0                    Fiction      406179.0\n","1           Juvenile Fiction       98825.0\n","2  Biography & Autobiography       90803.0\n","3                   Religion       32990.0\n","4                    History       32115.0\n","5                     Travel       28624.0\n","6        Young Adult Fiction       28508.0\n","7       Business & Economics       25425.0\n","8                 Philosophy       22215.0\n","9       Abnormalities, Human       20496.0"],"text/html":["\n","  <div id=\"df-e356e406-1eca-4c5a-96f1-880eb2a46182\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>categories</th>\n","      <th>ratingsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Fiction</td>\n","      <td>406179.0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Juvenile Fiction</td>\n","      <td>98825.0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Biography &amp; Autobiography</td>\n","      <td>90803.0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Religion</td>\n","      <td>32990.0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>History</td>\n","      <td>32115.0</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>Travel</td>\n","      <td>28624.0</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Young Adult Fiction</td>\n","      <td>28508.0</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Business &amp; Economics</td>\n","      <td>25425.0</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Philosophy</td>\n","      <td>22215.0</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>Abnormalities, Human</td>\n","      <td>20496.0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e356e406-1eca-4c5a-96f1-880eb2a46182')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-e356e406-1eca-4c5a-96f1-880eb2a46182 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-e356e406-1eca-4c5a-96f1-880eb2a46182');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":43}]},{"cell_type":"markdown","source":["#####*Year Published vs. Average Ratings Count*"],"metadata":{"id":"h41n33mQM8gi"}},{"cell_type":"markdown","source":["We want to see if there is a relationship between the year a book was published and the mean ratings count.\n","\n","*  We are only curious about modern books - let's filter on having been published in the 21st century.\n","*  Since the majority value of `ratingsCount` is 0, let's also filter on books having at least 1 rating.\n","\n"],"metadata":{"id":"tEt1WJxMPBse"}},{"cell_type":"code","source":["#TODO\n","books_2000s = cleaned_books_df[cleaned_books_df['published_year'].isin(['2000', '2001', '2002', '2003', '2004', '2005', '2006', '2007', '2008', '2009', '2010', '2011', '2012', '2013', '2014', '2015', '2016', '2017', '2018', '2019', '2020', '2021'])]\n","books_2000s = books_2000s[books_2000s['ratingsCount'] > 0]\n","avg_yearly_ratings = books_2000s[['published_year', 'ratingsCount']].groupby(by = 'published_year').mean().reset_index()"],"metadata":{"id":"bDEq1RFnRd69","executionInfo":{"status":"ok","timestamp":1671163117561,"user_tz":300,"elapsed":23,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":44,"outputs":[]},{"cell_type":"code","source":["import plotly.express as px\n","fig = px.line(avg_yearly_ratings, x= 'published_year', y='ratingsCount', title='Average Ratings Count for 21 year period',\n","              labels = {'published_year':'Year of Publish', 'ratingsCount': 'Average ratingsCount'}, markers=True)\n","fig.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":542},"id":"N0YvYrHYSVry","executionInfo":{"status":"ok","timestamp":1671163119829,"user_tz":300,"elapsed":2290,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"a7c5f0b9-6a72-42f8-e200-d9b6c134173c"},"execution_count":45,"outputs":[{"output_type":"display_data","data":{"text/html":["<html>\n","<head><meta charset=\"utf-8\" /></head>\n","<body>\n","    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n","        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"a4a004fd-3e35-4d07-93ef-3df62fcf1ffd\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"a4a004fd-3e35-4d07-93ef-3df62fcf1ffd\")) {                    Plotly.newPlot(                        \"a4a004fd-3e35-4d07-93ef-3df62fcf1ffd\",                        [{\"hovertemplate\":\"Year of Publish=%{x}<br>Average ratingsCount=%{y}<extra></extra>\",\"legendgroup\":\"\",\"line\":{\"color\":\"#636efa\",\"dash\":\"solid\"},\"marker\":{\"symbol\":\"circle\"},\"mode\":\"lines+markers\",\"name\":\"\",\"orientation\":\"v\",\"showlegend\":false,\"x\":[\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\"2020\",\"2021\"],\"xaxis\":\"x\",\"y\":[7.950305810397554,16.043785310734464,7.762753534111862,24.01252041371802,7.4099560761347,12.137452107279694,23.754927726675426,32.90680766688698,33.65818759936407,40.478391959799,35.555872291904215,23.51299656694458,36.680059523809526,33.179016004742145,60.60583446404342,45.45943482224248,29.381756756756758,46.241743725231174,12.924219910846954,25.77980132450331,20.248456790123456,14.114886731391586],\"yaxis\":\"y\",\"type\":\"scatter\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}},\"xaxis\":{\"anchor\":\"y\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Year of Publish\"}},\"yaxis\":{\"anchor\":\"x\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Average ratingsCount\"}},\"legend\":{\"tracegroupgap\":0},\"title\":{\"text\":\"Average Ratings Count for 21 year period\"}},                        {\"responsive\": true}                    ).then(function(){\n","                            \n","var gd = document.getElementById('a4a004fd-3e35-4d07-93ef-3df62fcf1ffd');\n","var x = new MutationObserver(function (mutations, observer) {{\n","        var display = window.getComputedStyle(gd).display;\n","        if (!display || display === 'none') {{\n","            console.log([gd, 'removed!']);\n","            Plotly.purge(gd);\n","            observer.disconnect();\n","        }}\n","}});\n","\n","// Listen for the removal of the full notebook cells\n","var notebookContainer = gd.closest('#notebook-container');\n","if (notebookContainer) {{\n","    x.observe(notebookContainer, {childList: true});\n","}}\n","\n","// Listen for the clearing of the current output cell\n","var outputEl = gd.closest('.output');\n","if (outputEl) {{\n","    x.observe(outputEl, {childList: true});\n","}}\n","\n","                        })                };                            </script>        </div>\n","</body>\n","</html>"]},"metadata":{}}]},{"cell_type":"markdown","source":["#####*Number of Authors and Categories vs. Ratings Count*"],"metadata":{"id":"O_csV7MTNSrV"}},{"cell_type":"markdown","source":["Do the number of authors of a given book or the number of categories of a given book relate to the average ratings count?\n","\n","*   Since it is not common for books to have more than 5 authors or categories we will limit on that.\n","*   Again we will examine books with a `ratingsCount` greater than 0.\n","\n"],"metadata":{"id":"OjkdZ0DJYsap"}},{"cell_type":"code","source":["num_books = cleaned_books_df[cleaned_books_df['num_authors'] <= 5]\n","num_books = num_books[num_books['num_categories'] <= 5]\n","num_books = num_books[num_books['ratingsCount'] > 0]\n","num_books = num_books[['ratingsCount', 'num_authors', 'num_categories']]"],"metadata":{"id":"1eZ7dlQSZanl","executionInfo":{"status":"ok","timestamp":1671163119830,"user_tz":300,"elapsed":16,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":46,"outputs":[]},{"cell_type":"code","source":["num_auths = num_books[['ratingsCount', 'num_authors']].groupby('num_authors').mean().reset_index().rename(columns = {'num_authors':'num'})\n","num_cats = num_books[['ratingsCount', 'num_categories']].groupby('num_categories').mean().reset_index().rename(columns = {'num_categories':'num'})\n","num_auths['Type'] = 'Number of Authors'\n","num_cats['Type'] = 'Number of Categories'"],"metadata":{"id":"lObbLbi3Zqny","executionInfo":{"status":"ok","timestamp":1671163119830,"user_tz":300,"elapsed":9,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":47,"outputs":[]},{"cell_type":"code","source":["plot_df = pd.concat([num_cats, num_auths])"],"metadata":{"id":"6kANeqvKaaAD","executionInfo":{"status":"ok","timestamp":1671163119834,"user_tz":300,"elapsed":12,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":48,"outputs":[]},{"cell_type":"code","source":["plot2 = sns.catplot(data = plot_df, x = 'num', y = 'ratingsCount', kind=\"bar\", col = 'Type', dodge = False)\n","plot2.set_axis_labels(\"\", \"Average Ratings Count\")\n","plot2.set_titles(\"{col_name} vs Average Ratings\")\n","plt.show(plot2)"],"metadata":{"id":"pio90Xsua5om","executionInfo":{"status":"ok","timestamp":1671163120502,"user_tz":300,"elapsed":679,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":355},"outputId":"f48f0f4b-e7a4-4299-f127-77d76a373816"},"execution_count":49,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 720x360 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","metadata":{"id":"CHFdRtQKLbti"},"source":["## Part 1.3: EDA with Reviews Data\n","Now that we're more familiar with the dataset, we will try and extract some information on which books seem to be popular by their total ratings, total number of reviews, percentile rankings, highly ranked authors etc. \n"]},{"cell_type":"markdown","source":["###1.3.1 Finding the popular books"],"metadata":{"id":"2UVthIxONbN9"}},{"cell_type":"markdown","metadata":{"id":"zHzBYzlDYoLx"},"source":["\n","\n","#### 1 Extracting total ratings\n","We now want to see which books were ranked highest. We can measure book popularity using the total number of reviews, dubbed reviewsCount, the book had.\n","\n","We will be using the `cleaned_reviews_df` dataframe in this section.\n","\n","\n","\n","\n"]},{"cell_type":"code","source":["cleaned_reviews_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":302},"id":"9kVrQ9wHV3Y-","executionInfo":{"status":"ok","timestamp":1671163120502,"user_tz":300,"elapsed":5,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"01269c54-1c86-413e-db10-dcdd349825ce"},"execution_count":50,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score  \n","0              positive  \n","1              positive  \n","2              positive  \n","3              positive  \n","4              positive  "],"text/html":["\n","  <div id=\"df-0f045dc0-bb14-4cdc-9651-e678d64e51d3\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0f045dc0-bb14-4cdc-9651-e678d64e51d3')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-0f045dc0-bb14-4cdc-9651-e678d64e51d3 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-0f045dc0-bb14-4cdc-9651-e678d64e51d3');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":50}]},{"cell_type":"code","execution_count":51,"metadata":{"id":"m2FvwI2CaA0X","executionInfo":{"status":"ok","timestamp":1671163121201,"user_tz":300,"elapsed":703,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"outputs":[],"source":["# Get number of reviews per title, dubbed reviewsCount, from reviews table --- should somewhat line up with ratingsCount\n","num_reviews = cleaned_reviews_df[['Title', 'review_score']].groupby('Title').count().reset_index().rename(columns={'review_score': 'reviewsCount'}).sort_values(by='reviewsCount',ascending=False)"]},{"cell_type":"markdown","source":["The number of reviews per title doesn't actually add up with the ratingsCount number from the books data. Let's do some more exploration on the books with most number of reviews as listed above\n","\n","\n","*   Extract the top ten books with the number of reviews\n","*   Find the books with the most number of **positive** reviews\n","\n"],"metadata":{"id":"qJO2D5mqOo-Q"}},{"cell_type":"code","source":["top10_reviewed = num_reviews.head(10)\n","top10_reviewed"],"metadata":{"id":"vJ1gZObqOn_I","executionInfo":{"status":"ok","timestamp":1671163121202,"user_tz":300,"elapsed":19,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"63473ceb-a38b-4f00-99cc-52063669400e"},"execution_count":52,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                        Title  reviewsCount\n","166772                             The Hobbit         22023\n","123880                    Pride and Prejudice         20371\n","17627                          Atlas Shrugged         12513\n","210376                      Wuthering Heights         10780\n","164670                              The Giver          7644\n","67537                      Great Expectations          7421\n","70370   Harry Potter and The Sorcerer's Stone          6796\n","113063                        Of Mice and Men          6728\n","26051                         Brave New World          6312\n","102195                      Mere Christianity          6053"],"text/html":["\n","  <div id=\"df-6f020e60-41f8-4381-bfb9-743e2997f178\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>reviewsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>166772</th>\n","      <td>The Hobbit</td>\n","      <td>22023</td>\n","    </tr>\n","    <tr>\n","      <th>123880</th>\n","      <td>Pride and Prejudice</td>\n","      <td>20371</td>\n","    </tr>\n","    <tr>\n","      <th>17627</th>\n","      <td>Atlas Shrugged</td>\n","      <td>12513</td>\n","    </tr>\n","    <tr>\n","      <th>210376</th>\n","      <td>Wuthering Heights</td>\n","      <td>10780</td>\n","    </tr>\n","    <tr>\n","      <th>164670</th>\n","      <td>The Giver</td>\n","      <td>7644</td>\n","    </tr>\n","    <tr>\n","      <th>67537</th>\n","      <td>Great Expectations</td>\n","      <td>7421</td>\n","    </tr>\n","    <tr>\n","      <th>70370</th>\n","      <td>Harry Potter and The Sorcerer's Stone</td>\n","      <td>6796</td>\n","    </tr>\n","    <tr>\n","      <th>113063</th>\n","      <td>Of Mice and Men</td>\n","      <td>6728</td>\n","    </tr>\n","    <tr>\n","      <th>26051</th>\n","      <td>Brave New World</td>\n","      <td>6312</td>\n","    </tr>\n","    <tr>\n","      <th>102195</th>\n","      <td>Mere Christianity</td>\n","      <td>6053</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6f020e60-41f8-4381-bfb9-743e2997f178')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-6f020e60-41f8-4381-bfb9-743e2997f178 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-6f020e60-41f8-4381-bfb9-743e2997f178');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":52}]},{"cell_type":"code","source":["positive_reviews_df = cleaned_reviews_df[cleaned_reviews_df['positive_review_score'] == 'positive']\n","num_positive_reviews = positive_reviews_df[['Title','review_score']].groupby('Title').count().reset_index().rename(columns={'review_score': 'pos_reviewsCount'}).sort_values(by='pos_reviewsCount',ascending=False)\n","\n","top10_positive_reviewed = num_positive_reviews.head(10)\n","top10_positive_reviewed"],"metadata":{"id":"70fCArJAl2a6","executionInfo":{"status":"ok","timestamp":1671163122227,"user_tz":300,"elapsed":1034,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"d83a2ecd-98f5-40dc-edba-bfba1f578138"},"execution_count":53,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                        Title  pos_reviewsCount\n","156262                             The Hobbit             20455\n","115905                    Pride and Prejudice             18013\n","16487                          Atlas Shrugged              9216\n","197291                      Wuthering Heights              7992\n","154251                              The Giver              6442\n","65804   Harry Potter and The Sorcerer's Stone              6395\n","63184                      Great Expectations              5747\n","105755                        Of Mice and Men              5746\n","95586                       Mere Christianity              5242\n","24385                         Brave New World              5144"],"text/html":["\n","  <div id=\"df-d27ecb0c-414f-468b-a409-af978c7cb2c4\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>pos_reviewsCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>156262</th>\n","      <td>The Hobbit</td>\n","      <td>20455</td>\n","    </tr>\n","    <tr>\n","      <th>115905</th>\n","      <td>Pride and Prejudice</td>\n","      <td>18013</td>\n","    </tr>\n","    <tr>\n","      <th>16487</th>\n","      <td>Atlas Shrugged</td>\n","      <td>9216</td>\n","    </tr>\n","    <tr>\n","      <th>197291</th>\n","      <td>Wuthering Heights</td>\n","      <td>7992</td>\n","    </tr>\n","    <tr>\n","      <th>154251</th>\n","      <td>The Giver</td>\n","      <td>6442</td>\n","    </tr>\n","    <tr>\n","      <th>65804</th>\n","      <td>Harry Potter and The Sorcerer's Stone</td>\n","      <td>6395</td>\n","    </tr>\n","    <tr>\n","      <th>63184</th>\n","      <td>Great Expectations</td>\n","      <td>5747</td>\n","    </tr>\n","    <tr>\n","      <th>105755</th>\n","      <td>Of Mice and Men</td>\n","      <td>5746</td>\n","    </tr>\n","    <tr>\n","      <th>95586</th>\n","      <td>Mere Christianity</td>\n","      <td>5242</td>\n","    </tr>\n","    <tr>\n","      <th>24385</th>\n","      <td>Brave New World</td>\n","      <td>5144</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d27ecb0c-414f-468b-a409-af978c7cb2c4')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-d27ecb0c-414f-468b-a409-af978c7cb2c4 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-d27ecb0c-414f-468b-a409-af978c7cb2c4');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":53}]},{"cell_type":"markdown","source":["\n","Let's find the percentage of positive reviews per title:\n","* Rank by number of reviews and percentage of positive reviews in descending order\n","\n"],"metadata":{"id":"KFa27CoYq_OL"}},{"cell_type":"code","source":["percent_pos_reviews = num_reviews.merge(num_positive_reviews,left_on='Title',right_on='Title', suffixes=[None,'_y']).rename(columns={'reviewsCount_y':'num_pos_reviews'})\n","percent_pos_reviews['percent_pos_reviews'] = percent_pos_reviews['pos_reviewsCount'] / percent_pos_reviews['reviewsCount']\n","percent_pos_reviews.sort_values(by=['reviewsCount','percent_pos_reviews'],ascending=[False,False]).head(10)"],"metadata":{"id":"GKuxZE11q8qi","executionInfo":{"status":"ok","timestamp":1671163122524,"user_tz":300,"elapsed":393,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"bc21511c-1674-4e00-a3ef-00fb13cafbe5"},"execution_count":54,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                   Title  reviewsCount  pos_reviewsCount  \\\n","0                             The Hobbit         22023             20455   \n","1                    Pride and Prejudice         20371             18013   \n","2                         Atlas Shrugged         12513              9216   \n","3                      Wuthering Heights         10780              7992   \n","4                              The Giver          7644              6442   \n","5                     Great Expectations          7421              5747   \n","6  Harry Potter and The Sorcerer's Stone          6796              6395   \n","7                        Of Mice and Men          6728              5746   \n","8                        Brave New World          6312              5144   \n","9                      Mere Christianity          6053              5242   \n","\n","   percent_pos_reviews  \n","0             0.928802  \n","1             0.884247  \n","2             0.736514  \n","3             0.741373  \n","4             0.842752  \n","5             0.774424  \n","6             0.940995  \n","7             0.854043  \n","8             0.814956  \n","9             0.866017  "],"text/html":["\n","  <div id=\"df-121d2d80-1914-4681-bfd9-daf2e8a7fd3d\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>reviewsCount</th>\n","      <th>pos_reviewsCount</th>\n","      <th>percent_pos_reviews</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>The Hobbit</td>\n","      <td>22023</td>\n","      <td>20455</td>\n","      <td>0.928802</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Pride and Prejudice</td>\n","      <td>20371</td>\n","      <td>18013</td>\n","      <td>0.884247</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Atlas Shrugged</td>\n","      <td>12513</td>\n","      <td>9216</td>\n","      <td>0.736514</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Wuthering Heights</td>\n","      <td>10780</td>\n","      <td>7992</td>\n","      <td>0.741373</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>The Giver</td>\n","      <td>7644</td>\n","      <td>6442</td>\n","      <td>0.842752</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>Great Expectations</td>\n","      <td>7421</td>\n","      <td>5747</td>\n","      <td>0.774424</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Harry Potter and The Sorcerer's Stone</td>\n","      <td>6796</td>\n","      <td>6395</td>\n","      <td>0.940995</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Of Mice and Men</td>\n","      <td>6728</td>\n","      <td>5746</td>\n","      <td>0.854043</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Brave New World</td>\n","      <td>6312</td>\n","      <td>5144</td>\n","      <td>0.814956</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>Mere Christianity</td>\n","      <td>6053</td>\n","      <td>5242</td>\n","      <td>0.866017</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-121d2d80-1914-4681-bfd9-daf2e8a7fd3d')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-121d2d80-1914-4681-bfd9-daf2e8a7fd3d button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-121d2d80-1914-4681-bfd9-daf2e8a7fd3d');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":54}]},{"cell_type":"markdown","source":["### Visualization\n","\n","Plot a multi-bar graph of number of reviews and the number of positive reviews per title of top 10 most rated books"],"metadata":{"id":"PPUmkNB7W2DB"}},{"cell_type":"code","source":["bar_reviews = percent_pos_reviews[['Title','reviewsCount','pos_reviewsCount']].head(10)\n","bar_reviews\n","# Plot grouped bar chart with number of reviews and number of positive reviews side by side\n","fig = px.bar(bar_reviews, x='Title', y=[\"reviewsCount\",'pos_reviewsCount'],\n","             barmode='group', labels={'value': 'Number of reviews', 'Title': 'Book Title','variable':'Split'},\n","             title = 'Distribution of Positive Reviews per Book for Top 10 Most Reviewed Books', height=600)\n","fig.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":617},"id":"XgYwcDvY2blh","executionInfo":{"status":"ok","timestamp":1671163123222,"user_tz":300,"elapsed":739,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"394f136b-93bd-4726-c895-9296797eb38d"},"execution_count":55,"outputs":[{"output_type":"display_data","data":{"text/html":["<html>\n","<head><meta charset=\"utf-8\" /></head>\n","<body>\n","    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n","        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"9772b0c5-4ea1-4e3b-a8ed-bc7ea6161804\" class=\"plotly-graph-div\" style=\"height:600px; width:100%;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"9772b0c5-4ea1-4e3b-a8ed-bc7ea6161804\")) {                    Plotly.newPlot(                        \"9772b0c5-4ea1-4e3b-a8ed-bc7ea6161804\",                        [{\"alignmentgroup\":\"True\",\"hovertemplate\":\"Split=reviewsCount<br>Book Title=%{x}<br>Number of reviews=%{y}<extra></extra>\",\"legendgroup\":\"reviewsCount\",\"marker\":{\"color\":\"#636efa\",\"pattern\":{\"shape\":\"\"}},\"name\":\"reviewsCount\",\"offsetgroup\":\"reviewsCount\",\"orientation\":\"v\",\"showlegend\":true,\"textposition\":\"auto\",\"x\":[\"The Hobbit\",\"Pride and Prejudice\",\"Atlas Shrugged\",\"Wuthering Heights\",\"The Giver\",\"Great Expectations\",\"Harry Potter and The Sorcerer's Stone\",\"Of Mice and Men\",\"Brave New World\",\"Mere Christianity\"],\"xaxis\":\"x\",\"y\":[22023,20371,12513,10780,7644,7421,6796,6728,6312,6053],\"yaxis\":\"y\",\"type\":\"bar\"},{\"alignmentgroup\":\"True\",\"hovertemplate\":\"Split=pos_reviewsCount<br>Book Title=%{x}<br>Number of reviews=%{y}<extra></extra>\",\"legendgroup\":\"pos_reviewsCount\",\"marker\":{\"color\":\"#EF553B\",\"pattern\":{\"shape\":\"\"}},\"name\":\"pos_reviewsCount\",\"offsetgroup\":\"pos_reviewsCount\",\"orientation\":\"v\",\"showlegend\":true,\"textposition\":\"auto\",\"x\":[\"The Hobbit\",\"Pride and Prejudice\",\"Atlas Shrugged\",\"Wuthering Heights\",\"The Giver\",\"Great Expectations\",\"Harry Potter and The Sorcerer's Stone\",\"Of Mice and Men\",\"Brave New World\",\"Mere Christianity\"],\"xaxis\":\"x\",\"y\":[20455,18013,9216,7992,6442,5747,6395,5746,5144,5242],\"yaxis\":\"y\",\"type\":\"bar\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}},\"xaxis\":{\"anchor\":\"y\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Book Title\"}},\"yaxis\":{\"anchor\":\"x\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Number of reviews\"}},\"legend\":{\"title\":{\"text\":\"Split\"},\"tracegroupgap\":0},\"title\":{\"text\":\"Distribution of Positive Reviews per Book for Top 10 Most Reviewed Books\"},\"barmode\":\"group\",\"height\":600},                        {\"responsive\": true}                    ).then(function(){\n","                            \n","var gd = document.getElementById('9772b0c5-4ea1-4e3b-a8ed-bc7ea6161804');\n","var x = new MutationObserver(function (mutations, observer) {{\n","        var display = window.getComputedStyle(gd).display;\n","        if (!display || display === 'none') {{\n","            console.log([gd, 'removed!']);\n","            Plotly.purge(gd);\n","            observer.disconnect();\n","        }}\n","}});\n","\n","// Listen for the removal of the full notebook cells\n","var notebookContainer = gd.closest('#notebook-container');\n","if (notebookContainer) {{\n","    x.observe(notebookContainer, {childList: true});\n","}}\n","\n","// Listen for the clearing of the current output cell\n","var outputEl = gd.closest('.output');\n","if (outputEl) {{\n","    x.observe(outputEl, {childList: true});\n","}}\n","\n","                        })                };                            </script>        </div>\n","</body>\n","</html>"]},"metadata":{}}]},{"cell_type":"markdown","metadata":{"id":"jkoazKrGz-eP"},"source":["#Part 1.4: Combining the Data \n","\n","Since our data is spread out across two primary dataframes, books and reviews, it might be useful to aggregate the data for further use.\n","\n","In doing so, some more EDA can be done with the combined fields to reveal more interesting trends and statistics about the relation between the two dataframes."]},{"cell_type":"code","source":["#TODO\n","cleaned_books_df.head()"],"metadata":{"id":"LVfO-kdKSI1z","executionInfo":{"status":"ok","timestamp":1671163123225,"user_tz":300,"elapsed":79,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":302},"outputId":"99732b80-e4c0-48b8-e41c-2ddcd6215bb5"},"execution_count":56,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title  \\\n","0                           Dr. Seuss: American Icon   \n","1                      Whispers of the Wicked Saints   \n","2  The Church of Christ: A Biblical Ecclesiology ...   \n","3                           Saint Hyacinth of Poland   \n","4  Rising Sons and Daughters: Life Among Japan's ...   \n","\n","                                         description  \\\n","0  Philip Nel takes a fascinating look into the k...   \n","1  Julia Thomas finds her life spinning out of co...   \n","2  In The Church of Christ: A Biblical Ecclesiolo...   \n","3  The story for children 10 and up of St. Hyacin...   \n","4  Wardell recalls his experience as a foreign st...   \n","\n","                    authors              publisher  \\\n","0            ['Philip Nel']                ACBlack   \n","1       ['Veronica Haddon']              iUniverse   \n","2      ['Everett Ferguson']  WmBEerdmansPublishing   \n","3  ['Mary Fabyan Windeatt']            TanBooksPub   \n","4        ['Steven Wardell']      PlymptonPressIntl   \n","\n","                      categories  ratingsCount published_year  num_authors  \\\n","0  ['Biography & Autobiography']           0.0           2005            1   \n","1                    ['Fiction']           0.0           2005            1   \n","2                   ['Religion']           5.0           1996            1   \n","3  ['Biography & Autobiography']           0.0           2009            1   \n","4             ['Social Science']           0.0           1995            1   \n","\n","   num_categories  \n","0               1  \n","1               1  \n","2               1  \n","3               1  \n","4               1  "],"text/html":["\n","  <div id=\"df-c4d8f9d2-5171-4a9a-a528-6f4901a34ba6\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>description</th>\n","      <th>authors</th>\n","      <th>publisher</th>\n","      <th>categories</th>\n","      <th>ratingsCount</th>\n","      <th>published_year</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>Philip Nel takes a fascinating look into the k...</td>\n","      <td>['Philip Nel']</td>\n","      <td>ACBlack</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>0.0</td>\n","      <td>2005</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Whispers of the Wicked Saints</td>\n","      <td>Julia Thomas finds her life spinning out of co...</td>\n","      <td>['Veronica Haddon']</td>\n","      <td>iUniverse</td>\n","      <td>['Fiction']</td>\n","      <td>0.0</td>\n","      <td>2005</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>The Church of Christ: A Biblical Ecclesiology ...</td>\n","      <td>In The Church of Christ: A Biblical Ecclesiolo...</td>\n","      <td>['Everett Ferguson']</td>\n","      <td>WmBEerdmansPublishing</td>\n","      <td>['Religion']</td>\n","      <td>5.0</td>\n","      <td>1996</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Saint Hyacinth of Poland</td>\n","      <td>The story for children 10 and up of St. Hyacin...</td>\n","      <td>['Mary Fabyan Windeatt']</td>\n","      <td>TanBooksPub</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>0.0</td>\n","      <td>2009</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Rising Sons and Daughters: Life Among Japan's ...</td>\n","      <td>Wardell recalls his experience as a foreign st...</td>\n","      <td>['Steven Wardell']</td>\n","      <td>PlymptonPressIntl</td>\n","      <td>['Social Science']</td>\n","      <td>0.0</td>\n","      <td>1995</td>\n","      <td>1</td>\n","      <td>1</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c4d8f9d2-5171-4a9a-a528-6f4901a34ba6')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-c4d8f9d2-5171-4a9a-a528-6f4901a34ba6 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-c4d8f9d2-5171-4a9a-a528-6f4901a34ba6');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":56}]},{"cell_type":"code","source":["#TODO\n","cleaned_reviews_df.head()"],"metadata":{"id":"ZANTQVQ6SJFB","executionInfo":{"status":"ok","timestamp":1671163123233,"user_tz":300,"elapsed":54,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":302},"outputId":"e007d27e-dffe-47f9-d9f9-45d45cb84cd6"},"execution_count":57,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score  \n","0              positive  \n","1              positive  \n","2              positive  \n","3              positive  \n","4              positive  "],"text/html":["\n","  <div id=\"df-22603dba-adf0-4c1d-812e-80b09e952431\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-22603dba-adf0-4c1d-812e-80b09e952431')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-22603dba-adf0-4c1d-812e-80b09e952431 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-22603dba-adf0-4c1d-812e-80b09e952431');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":57}]},{"cell_type":"markdown","source":["Observing the books and reviews dataframe, relevant fields that we would like to join for our analysis include: \n","- From Books dataframe: Title, authors, publisher and categories \n","- From Reviews dataframe: review_summary, review_score, positive_review_score"],"metadata":{"id":"FNr8_mTLv2Cj"}},{"cell_type":"code","source":["#TODO\n","combined_df = cleaned_books_df.merge(cleaned_reviews_df, on = 'Title', how='right',indicator=True)[['Title','publisher','published_year','categories','review_summary','review_score','positive_review_score']]"],"metadata":{"id":"Dt9gCfEvSJNT","executionInfo":{"status":"ok","timestamp":1671163126788,"user_tz":300,"elapsed":3598,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":58,"outputs":[]},{"cell_type":"code","source":["combined_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"jmH2KNTRxqA0","executionInfo":{"status":"ok","timestamp":1671163126789,"user_tz":300,"elapsed":151,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"d1775639-b845-4613-a56b-9340cb3cd131"},"execution_count":59,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title publisher published_year  \\\n","0  Its Only Art If Its Well Hung!       NaN            NaN   \n","1        Dr. Seuss: American Icon   ACBlack           2005   \n","2        Dr. Seuss: American Icon   ACBlack           2005   \n","3        Dr. Seuss: American Icon   ACBlack           2005   \n","4        Dr. Seuss: American Icon   ACBlack           2005   \n","\n","                      categories  \\\n","0                            NaN   \n","1  ['Biography & Autobiography']   \n","2  ['Biography & Autobiography']   \n","3  ['Biography & Autobiography']   \n","4  ['Biography & Autobiography']   \n","\n","                                    review_summary  review_score  \\\n","0           Nice collection of Julie Strain images           4.0   \n","1                                Really Enjoyed It           5.0   \n","2  Essential for every personal and Public Library           5.0   \n","3  Phlip Nel gives silly Seuss a serious treatment           4.0   \n","4                           Good academic overview           4.0   \n","\n","  positive_review_score  \n","0              positive  \n","1              positive  \n","2              positive  \n","3              positive  \n","4              positive  "],"text/html":["\n","  <div id=\"df-8ff242ae-e49f-46c7-967b-c830b2d499fb\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>publisher</th>\n","      <th>published_year</th>\n","      <th>categories</th>\n","      <th>review_summary</th>\n","      <th>review_score</th>\n","      <th>positive_review_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>NaN</td>\n","      <td>NaN</td>\n","      <td>NaN</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>4.0</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>ACBlack</td>\n","      <td>2005</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>Really Enjoyed It</td>\n","      <td>5.0</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>ACBlack</td>\n","      <td>2005</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>5.0</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>ACBlack</td>\n","      <td>2005</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>4.0</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>ACBlack</td>\n","      <td>2005</td>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>Good academic overview</td>\n","      <td>4.0</td>\n","      <td>positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8ff242ae-e49f-46c7-967b-c830b2d499fb')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-8ff242ae-e49f-46c7-967b-c830b2d499fb button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-8ff242ae-e49f-46c7-967b-c830b2d499fb');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":59}]},{"cell_type":"markdown","source":["Some interesting trends we would like to extract:\n","\n","- Top 10 **publishers** by number of reviews\n","- Top 10 **categories** by number of reviews\n","- Top 10 **authors** by number of reviews"],"metadata":{"id":"dnc0iBN8wmKI"}},{"cell_type":"markdown","source":["*Top 10 publishers by reviewsCount*"],"metadata":{"id":"DGD1aNMroSQD"}},{"cell_type":"code","source":["# Top 8 publishers by reviewsCount\n","\n","top10_publishers = combined_df[['publisher','Title']].groupby('publisher').count().rename(columns={'Title':'publisherCount'}).sort_values(by='publisherCount',ascending=False).reset_index().head(10)\n","top10_publishers"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"c-VBuKRsj9X9","executionInfo":{"status":"ok","timestamp":1671163128095,"user_tz":300,"elapsed":1391,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"d892ebf5-23af-41de-de07-366211528c84"},"execution_count":60,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                 publisher  publisherCount\n","0                  Penguin          136767\n","1         SimonandSchuster          117631\n","2            HarperCollins          113318\n","3              RandomHouse           46171\n","4                  Vintage           41527\n","5       CourierCorporation           37577\n","6             MarinerBooks           30142\n","7                Macmillan           29179\n","8  HoughtonMifflinHarcourt           28031\n","9                PenguinUK           26795"],"text/html":["\n","  <div id=\"df-bb1d9520-ba01-4a4a-8ea7-f1a81e483bd3\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>publisher</th>\n","      <th>publisherCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Penguin</td>\n","      <td>136767</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>SimonandSchuster</td>\n","      <td>117631</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>HarperCollins</td>\n","      <td>113318</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>RandomHouse</td>\n","      <td>46171</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Vintage</td>\n","      <td>41527</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>CourierCorporation</td>\n","      <td>37577</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>MarinerBooks</td>\n","      <td>30142</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Macmillan</td>\n","      <td>29179</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>HoughtonMifflinHarcourt</td>\n","      <td>28031</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>PenguinUK</td>\n","      <td>26795</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-bb1d9520-ba01-4a4a-8ea7-f1a81e483bd3')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-bb1d9520-ba01-4a4a-8ea7-f1a81e483bd3 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-bb1d9520-ba01-4a4a-8ea7-f1a81e483bd3');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":60}]},{"cell_type":"markdown","source":["One would expect that the longer a book has been published, the more reviews it probably amasses. Therefore, books with an older publishing date would have more reviews than newer published material. Let's see if this trend holds true in our dataset."],"metadata":{"id":"jL7-nMbSyaKP"}},{"cell_type":"code","source":["reviews_per_year = combined_df[['publisher','published_year']].groupby('published_year').count().rename(columns={'publisher':'num_reviews'}).sort_values(by='published_year',ascending=True).reset_index()\n","\n","# Truncate for books published in the 21st century\n","reviews_per_year = reviews_per_year[reviews_per_year['published_year'] >= '2000']"],"metadata":{"id":"VG7Ec86-ysPL","executionInfo":{"status":"ok","timestamp":1671163128097,"user_tz":300,"elapsed":96,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":61,"outputs":[]},{"cell_type":"code","source":["import plotly.express as px\n","fig = px.line(reviews_per_year, x= 'published_year', y='num_reviews', title='Number of Amazon Book Reviews for Books in a Given Year of Publication',\n","              labels = {'published_year':'Year of Publication', 'num_reviews': 'Number of reviews'}, markers=True)\n","fig.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":542},"id":"DlHJyD23yXdF","executionInfo":{"status":"ok","timestamp":1671163128101,"user_tz":300,"elapsed":74,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"6c27738a-7554-42a2-86c7-6d25bec08c94"},"execution_count":62,"outputs":[{"output_type":"display_data","data":{"text/html":["<html>\n","<head><meta charset=\"utf-8\" /></head>\n","<body>\n","    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n","        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"7380750b-fb61-4c8b-a240-c0efd421dc57\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"7380750b-fb61-4c8b-a240-c0efd421dc57\")) {                    Plotly.newPlot(                        \"7380750b-fb61-4c8b-a240-c0efd421dc57\",                        [{\"hovertemplate\":\"Year of Publication=%{x}<br>Number of reviews=%{y}<extra></extra>\",\"legendgroup\":\"\",\"line\":{\"color\":\"#636efa\",\"dash\":\"solid\"},\"marker\":{\"symbol\":\"circle\"},\"mode\":\"lines+markers\",\"name\":\"\",\"orientation\":\"v\",\"showlegend\":false,\"x\":[\"2000\",\"2001\",\"2002\",\"2003\",\"2004\",\"2005\",\"2006\",\"2007\",\"2008\",\"2009\",\"2010\",\"2011\",\"2012\",\"2013\",\"2014\",\"2015\",\"2016\",\"2017\",\"2018\",\"2019\",\"2020\",\"2021\",\"2022\",\"2023\",\"2030\"],\"xaxis\":\"x\",\"y\":[58662,76516,69950,87985,74495,87820,69142,78345,66065,113633,106928,97146,136561,75782,87841,56647,63102,40615,45327,39133,36672,41777,12740,60,5],\"yaxis\":\"y\",\"type\":\"scatter\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}},\"xaxis\":{\"anchor\":\"y\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Year of Publication\"}},\"yaxis\":{\"anchor\":\"x\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"Number of reviews\"}},\"legend\":{\"tracegroupgap\":0},\"title\":{\"text\":\"Number of Amazon Book Reviews for Books in a Given Year of Publication\"}},                        {\"responsive\": true}                    ).then(function(){\n","                            \n","var gd = document.getElementById('7380750b-fb61-4c8b-a240-c0efd421dc57');\n","var x = new MutationObserver(function (mutations, observer) {{\n","        var display = window.getComputedStyle(gd).display;\n","        if (!display || display === 'none') {{\n","            console.log([gd, 'removed!']);\n","            Plotly.purge(gd);\n","            observer.disconnect();\n","        }}\n","}});\n","\n","// Listen for the removal of the full notebook cells\n","var notebookContainer = gd.closest('#notebook-container');\n","if (notebookContainer) {{\n","    x.observe(notebookContainer, {childList: true});\n","}}\n","\n","// Listen for the clearing of the current output cell\n","var outputEl = gd.closest('.output');\n","if (outputEl) {{\n","    x.observe(outputEl, {childList: true});\n","}}\n","\n","                        })                };                            </script>        </div>\n","</body>\n","</html>"]},"metadata":{}}]},{"cell_type":"markdown","source":["Books published in the mid 2000s (2008-2014) seem to be very popular (or at least have a lot of reviews about them) despite seeing a somehwat constant trend in number of reviews over the years in the years prior\n","\n","The top 4 books by reviewsCount lead by a sizeable margin so if we let's observe the trend again if we remove by "],"metadata":{"id":"RmB9ch6G1cyc"}},{"cell_type":"markdown","source":["*Top 10 categories by reviewsCount*"],"metadata":{"id":"8P6z9nszoYqf"}},{"cell_type":"code","source":["top10_cats_df = combined_df[['categories', 'Title']].groupby(by='categories').count().sort_values(by='Title', ascending = False).head(10).reset_index()\n","top10_cats_df['categories'] = top10_cats_df['categories'].apply(lambda x: x[2:len(x) -2])\n","top10_cats_df"],"metadata":{"id":"1DMI1GrDSJSo","executionInfo":{"status":"ok","timestamp":1671163128103,"user_tz":300,"elapsed":58,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":363},"outputId":"ab7610e1-7d21-4a31-9779-2bb2575e4069"},"execution_count":63,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                  categories   Title\n","0                    Fiction  797054\n","1           Juvenile Fiction  198106\n","2  Biography & Autobiography   98754\n","3                   Religion   89665\n","4                    History   83688\n","5       Business & Economics   61060\n","6                  Computers   38611\n","7        Young Adult Fiction   28507\n","8             Social Science   27849\n","9                  Self-Help   27628"],"text/html":["\n","  <div id=\"df-f722e800-cb5a-4560-98f4-8ace15568e28\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>categories</th>\n","      <th>Title</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Fiction</td>\n","      <td>797054</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Juvenile Fiction</td>\n","      <td>198106</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Biography &amp; Autobiography</td>\n","      <td>98754</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Religion</td>\n","      <td>89665</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>History</td>\n","      <td>83688</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>Business &amp; Economics</td>\n","      <td>61060</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>Computers</td>\n","      <td>38611</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>Young Adult Fiction</td>\n","      <td>28507</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>Social Science</td>\n","      <td>27849</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>Self-Help</td>\n","      <td>27628</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f722e800-cb5a-4560-98f4-8ace15568e28')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-f722e800-cb5a-4560-98f4-8ace15568e28 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-f722e800-cb5a-4560-98f4-8ace15568e28');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":63}]},{"cell_type":"markdown","source":["*Top 10 authors by reviewsCount*"],"metadata":{"id":"K_7M2GJZ5niv"}},{"cell_type":"code","source":["combined_df_author = cleaned_books_df.merge(cleaned_reviews_df, on = 'Title', how='left',indicator=True)[['Title', 'authors','publisher','categories','review_summary','review_score','positive_review_score','_merge']]\n","top10_authors_df = combined_df_author[['authors','Title']].groupby('authors').count().rename(columns={'Title':'authorCount'}).sort_values(by='authorCount',ascending=False).reset_index().head(10)\n","top10_authors_df['authors'] = top10_authors_df['authors'].apply(lambda x: x[2:len(x) -2])\n","\n","top10_authors_df"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"A_W7zPS16Wk-","executionInfo":{"status":"ok","timestamp":1671163131414,"user_tz":300,"elapsed":3360,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"2a2e32a4-0b89-4079-eb76-ada62be1e867"},"execution_count":64,"outputs":[{"output_type":"execute_result","data":{"text/plain":["               authors  authorCount\n","0     J. R. R. Tolkien        33906\n","1          Jane Austen        29866\n","2      Charles Dickens        15978\n","3       John Steinbeck        12203\n","4        Kurt Vonnegut        12093\n","5           Harper Lee        12013\n","6          C. S. Lewis        11800\n","7  F. Scott Fitzgerald        10535\n","8        George Orwell        10098\n","9        J. K. Rowling         9942"],"text/html":["\n","  <div id=\"df-b2b1232b-c086-43e2-9f27-780294be51eb\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>authors</th>\n","      <th>authorCount</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>J. R. R. Tolkien</td>\n","      <td>33906</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Jane Austen</td>\n","      <td>29866</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Charles Dickens</td>\n","      <td>15978</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>John Steinbeck</td>\n","      <td>12203</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Kurt Vonnegut</td>\n","      <td>12093</td>\n","    </tr>\n","    <tr>\n","      <th>5</th>\n","      <td>Harper Lee</td>\n","      <td>12013</td>\n","    </tr>\n","    <tr>\n","      <th>6</th>\n","      <td>C. S. Lewis</td>\n","      <td>11800</td>\n","    </tr>\n","    <tr>\n","      <th>7</th>\n","      <td>F. Scott Fitzgerald</td>\n","      <td>10535</td>\n","    </tr>\n","    <tr>\n","      <th>8</th>\n","      <td>George Orwell</td>\n","      <td>10098</td>\n","    </tr>\n","    <tr>\n","      <th>9</th>\n","      <td>J. K. Rowling</td>\n","      <td>9942</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b2b1232b-c086-43e2-9f27-780294be51eb')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-b2b1232b-c086-43e2-9f27-780294be51eb button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-b2b1232b-c086-43e2-9f27-780294be51eb');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":64}]},{"cell_type":"markdown","metadata":{"id":"QbNUhlz9ftHI"},"source":["# Part 2: Working with Text Data\n","\n","We will now delve into text-based analysis. Although textual data can be and is complex, it can also be used to generate extremely interpretable results which makes it a valuable asset in data analysis as we can draw interesting insights from it. \n","\n","We now attempt to answer:\n","**According to the `cleaned_books_df` dataframe, what do the reviews for the top rated books look like?**\n","\n","<br>\n","<center><img src = \"https://cdn.xxl.thumbs.canstockphoto.com/word-clouds-iii-a-set-of-26-classic-style-comic-word-clouds-and-symbols-for-voice-thought-eps-vector_csp4426414.jpg\" width= \"400\" align =\"center\"/></center>\n","<br>\n","\n"]},{"cell_type":"markdown","metadata":{"id":"_h2xGHMKeIW8"},"source":["###2.1 Text cleaning\n","We first want to clean the text first using regular expressions (regex).\n","\n"]},{"cell_type":"code","source":["cleaned_reviews_df['review_text']"],"metadata":{"id":"6KrCqYH5eSWe","executionInfo":{"status":"ok","timestamp":1671163131416,"user_tz":300,"elapsed":42,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"a513575e-3105-422a-910a-ffb7f9e0379c"},"execution_count":65,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0          This is only for Julie Strain fans. It's a col...\n","1          I don't care much for Dr. Seuss but after read...\n","2          If people become the books they read and if \"t...\n","3          Theodore Seuss Geisel (1904-1991), aka &quot;D...\n","4          Philip Nel - Dr. Seuss: American IconThis is b...\n","                                 ...                        \n","2999995    This is an extremely difficult book to digest,...\n","2999996    This is pretty interesting. Collingwood seems ...\n","2999997    This is a good book but very esoteric. \"What i...\n","2999998    My daughter, a freshman at Indiana University,...\n","2999999    The guy has a few good ideas but, reader, bewa...\n","Name: review_text, Length: 2999746, dtype: object"]},"metadata":{},"execution_count":65}]},{"cell_type":"markdown","metadata":{"id":"RNwLCBeTOLy3"},"source":["###2.2 Text tokenization \n","\n","We are going to split the contents of in the Reviews column into a list of words. We will use the **nltk** library, which contains an extensive set of tools for text processing. \n","\n","Functions we've found useful in the past include: \n","- `nltk.word_tokenize()`: a function used to tokenize text\n","- `nltk.corpus.stopwords`: a list of commonly used words such as \"a\", \"an\",\"in\" that are often ignored in text analysis\n","\n","\n","\n","**TODO:**\n","- Use **nltk.corpus.stopwords** to create a set containing the most common English stopwords.\n","- Implement a function that  takes in a string and:\n","\n","1. Tokenize text such as review_summary per title\n","2. Find most common words per title\n","3. Produce a word cloud for top tokens\n","4. Calculate a sentiment score based on the review itself and compare to the actual review score the user gives"]},{"cell_type":"code","source":["cleaned_reviews_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":302},"id":"mrvjRajtZfWS","executionInfo":{"status":"ok","timestamp":1671163131417,"user_tz":300,"elapsed":24,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"427a8fac-f678-4d54-cf8a-4e48aa5d0b53"},"execution_count":66,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score  \n","0              positive  \n","1              positive  \n","2              positive  \n","3              positive  \n","4              positive  "],"text/html":["\n","  <div id=\"df-864586d3-4c94-4d62-bc9b-ecc92dc904ec\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-864586d3-4c94-4d62-bc9b-ecc92dc904ec')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-864586d3-4c94-4d62-bc9b-ecc92dc904ec button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-864586d3-4c94-4d62-bc9b-ecc92dc904ec');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":66}]},{"cell_type":"code","source":["# Necessary wall of imports\n","import nltk\n","nltk.download('punkt')\n","nltk.__version__"],"metadata":{"id":"NFfezb4-cFUE","executionInfo":{"status":"ok","timestamp":1671163131588,"user_tz":300,"elapsed":190,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/","height":72},"outputId":"083ab284-57bb-42a3-f28e-bee7d8b9f658"},"execution_count":67,"outputs":[{"output_type":"stream","name":"stderr","text":["[nltk_data] Downloading package punkt to /root/nltk_data...\n","[nltk_data]   Package punkt is already up-to-date!\n"]},{"output_type":"execute_result","data":{"text/plain":["'3.7'"],"application/vnd.google.colaboratory.intrinsic+json":{"type":"string"}},"metadata":{},"execution_count":67}]},{"cell_type":"code","source":["from nltk.corpus import stopwords\n","nltk.download('stopwords')\n","stopwords = set(stopwords.words('english'))"],"metadata":{"id":"t7NXd_eQXF3N","executionInfo":{"status":"ok","timestamp":1671163131589,"user_tz":300,"elapsed":14,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"fd0fa9c7-7e0d-4eb7-cf56-325c7bc8b6c9"},"execution_count":68,"outputs":[{"output_type":"stream","name":"stderr","text":["[nltk_data] Downloading package stopwords to /root/nltk_data...\n","[nltk_data]   Package stopwords is already up-to-date!\n"]}]},{"cell_type":"code","source":["from nltk.corpus.reader.tagged import word_tokenize\n","# Function to tokenize review text\n","def tokenize_content(content):\n","  words_cleaned = []\n","  words = word_tokenize(content)\n","  words = [word.lower() for word in words if word.isalpha()]\n","\n","  for word in words:\n","    if word not in stopwords:\n","      words_cleaned.append(word)\n","  return words_cleaned"],"metadata":{"id":"NsUQLj7oXKVa","executionInfo":{"status":"ok","timestamp":1671163131589,"user_tz":300,"elapsed":7,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":69,"outputs":[]},{"cell_type":"code","source":["review_tokens_df = cleaned_reviews_df"],"metadata":{"id":"8KsrlRikQadM","executionInfo":{"status":"ok","timestamp":1671163131589,"user_tz":300,"elapsed":7,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}}},"execution_count":70,"outputs":[]},{"cell_type":"code","source":["review_tokens_df = cleaned_reviews_df\n","# Tokenize review summaries per title and store in a list. WARNING: Code takes about 5 minutes to run even for the summaries (unsure how much for full text)\n","review_tokens_df['review_summary_tokens'] = review_tokens_df['review_summary'].apply(tokenize_content)\n","review_tokens_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":337},"id":"6yYkTUOwMAjv","executionInfo":{"status":"ok","timestamp":1671163366680,"user_tz":300,"elapsed":235098,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"55dc4abd-6500-4fb8-98e9-d2b744065bb7"},"execution_count":71,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score                              review_summary_tokens  \n","0              positive          [nice, collection, julie, strain, images]  \n","1              positive                                  [really, enjoyed]  \n","2              positive      [essential, every, personal, public, library]  \n","3              positive  [phlip, nel, gives, silly, seuss, serious, tre...  \n","4              positive                         [good, academic, overview]  "],"text/html":["\n","  <div id=\"df-51050d9b-f208-4081-a15d-f18cffa163f7\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","      <th>review_summary_tokens</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[nice, collection, julie, strain, images]</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[really, enjoyed]</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","      <td>[essential, every, personal, public, library]</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[phlip, nel, gives, silly, seuss, serious, tre...</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[good, academic, overview]</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-51050d9b-f208-4081-a15d-f18cffa163f7')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-51050d9b-f208-4081-a15d-f18cffa163f7 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-51050d9b-f208-4081-a15d-f18cffa163f7');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":71}]},{"cell_type":"code","source":["# Find most common words per title (might remove this section, review summary is too short for this analysis most likely)\n","from collections import Counter\n","review_tokens_df['most_common_words'] = review_tokens_df['review_summary_tokens'].apply(lambda x: Counter(x).most_common(5))\n","review_tokens_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":424},"id":"T8s3Yj2MOeq7","executionInfo":{"status":"ok","timestamp":1671163380059,"user_tz":300,"elapsed":13482,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"153bf2ba-5e1b-42fd-9860-264d619d68da"},"execution_count":72,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score                              review_summary_tokens  \\\n","0              positive          [nice, collection, julie, strain, images]   \n","1              positive                                  [really, enjoyed]   \n","2              positive      [essential, every, personal, public, library]   \n","3              positive  [phlip, nel, gives, silly, seuss, serious, tre...   \n","4              positive                         [good, academic, overview]   \n","\n","                                   most_common_words  \n","0  [(nice, 1), (collection, 1), (julie, 1), (stra...  \n","1                        [(really, 1), (enjoyed, 1)]  \n","2  [(essential, 1), (every, 1), (personal, 1), (p...  \n","3  [(phlip, 1), (nel, 1), (gives, 1), (silly, 1),...  \n","4          [(good, 1), (academic, 1), (overview, 1)]  "],"text/html":["\n","  <div id=\"df-c1cffa31-68fd-4d19-9cc6-84e16879b6b4\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","      <th>review_summary_tokens</th>\n","      <th>most_common_words</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[nice, collection, julie, strain, images]</td>\n","      <td>[(nice, 1), (collection, 1), (julie, 1), (stra...</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[really, enjoyed]</td>\n","      <td>[(really, 1), (enjoyed, 1)]</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","      <td>[essential, every, personal, public, library]</td>\n","      <td>[(essential, 1), (every, 1), (personal, 1), (p...</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[phlip, nel, gives, silly, seuss, serious, tre...</td>\n","      <td>[(phlip, 1), (nel, 1), (gives, 1), (silly, 1),...</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[good, academic, overview]</td>\n","      <td>[(good, 1), (academic, 1), (overview, 1)]</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c1cffa31-68fd-4d19-9cc6-84e16879b6b4')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-c1cffa31-68fd-4d19-9cc6-84e16879b6b4 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-c1cffa31-68fd-4d19-9cc6-84e16879b6b4');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":72}]},{"cell_type":"code","source":["# Compute Afinn score on review_summary, normalize to same scale as review score and compare the two\n","!pip install afinn"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"2GrttgpOQilp","executionInfo":{"status":"ok","timestamp":1671163383373,"user_tz":300,"elapsed":3326,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"4d34475a-f887-42cb-8c18-7e413693ce8d"},"execution_count":73,"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Requirement already satisfied: afinn in /usr/local/lib/python3.8/dist-packages (0.1)\n"]}]},{"cell_type":"code","source":["from afinn import Afinn\n","afinn = Afinn()\n","# WARNING: this cell takes about 8 minutes to run. Couldn't get line to run with full review text\n","review_tokens_df['sentiment_score'] = review_tokens_df['review_summary'].apply(afinn.score)\n","review_tokens_df.head()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":531},"id":"GTRQ7vOUXVzD","executionInfo":{"status":"ok","timestamp":1671163613348,"user_tz":300,"elapsed":229981,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"0af25950-5968-45c0-85e6-73941fddc80a"},"execution_count":74,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                            Title         User_id review_helpfulness_rating  \\\n","0  Its Only Art If Its Well Hung!   AVCGYZL8FQQTD                       7/7   \n","1        Dr. Seuss: American Icon  A30TK6U7DNS82R                     10/10   \n","2        Dr. Seuss: American Icon  A3UH4UZ4RSVO82                     10/11   \n","3        Dr. Seuss: American Icon  A2MVUWT453QH61                       7/7   \n","4        Dr. Seuss: American Icon  A22X4XUPKF66MR                       3/3   \n","\n","   review_score                                   review_summary  \\\n","0           4.0           Nice collection of Julie Strain images   \n","1           5.0                                Really Enjoyed It   \n","2           5.0  Essential for every personal and Public Library   \n","3           4.0  Phlip Nel gives silly Seuss a serious treatment   \n","4           4.0                           Good academic overview   \n","\n","                                         review_text  review_percent_helpful  \\\n","0  This is only for Julie Strain fans. It's a col...                  100.00   \n","1  I don't care much for Dr. Seuss but after read...                  100.00   \n","2  If people become the books they read and if \"t...                   90.91   \n","3  Theodore Seuss Geisel (1904-1991), aka &quot;D...                  100.00   \n","4  Philip Nel - Dr. Seuss: American IconThis is b...                  100.00   \n","\n","  positive_review_score                              review_summary_tokens  \\\n","0              positive          [nice, collection, julie, strain, images]   \n","1              positive                                  [really, enjoyed]   \n","2              positive      [essential, every, personal, public, library]   \n","3              positive  [phlip, nel, gives, silly, seuss, serious, tre...   \n","4              positive                         [good, academic, overview]   \n","\n","                                   most_common_words  sentiment_score  \n","0  [(nice, 1), (collection, 1), (julie, 1), (stra...              3.0  \n","1                        [(really, 1), (enjoyed, 1)]              2.0  \n","2  [(essential, 1), (every, 1), (personal, 1), (p...              0.0  \n","3  [(phlip, 1), (nel, 1), (gives, 1), (silly, 1),...             -1.0  \n","4          [(good, 1), (academic, 1), (overview, 1)]              3.0  "],"text/html":["\n","  <div id=\"df-2e9ec9ff-f6bb-4297-9f5d-b217ddfc810e\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>User_id</th>\n","      <th>review_helpfulness_rating</th>\n","      <th>review_score</th>\n","      <th>review_summary</th>\n","      <th>review_text</th>\n","      <th>review_percent_helpful</th>\n","      <th>positive_review_score</th>\n","      <th>review_summary_tokens</th>\n","      <th>most_common_words</th>\n","      <th>sentiment_score</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Its Only Art If Its Well Hung!</td>\n","      <td>AVCGYZL8FQQTD</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Nice collection of Julie Strain images</td>\n","      <td>This is only for Julie Strain fans. It's a col...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[nice, collection, julie, strain, images]</td>\n","      <td>[(nice, 1), (collection, 1), (julie, 1), (stra...</td>\n","      <td>3.0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A30TK6U7DNS82R</td>\n","      <td>10/10</td>\n","      <td>5.0</td>\n","      <td>Really Enjoyed It</td>\n","      <td>I don't care much for Dr. Seuss but after read...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[really, enjoyed]</td>\n","      <td>[(really, 1), (enjoyed, 1)]</td>\n","      <td>2.0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A3UH4UZ4RSVO82</td>\n","      <td>10/11</td>\n","      <td>5.0</td>\n","      <td>Essential for every personal and Public Library</td>\n","      <td>If people become the books they read and if \"t...</td>\n","      <td>90.91</td>\n","      <td>positive</td>\n","      <td>[essential, every, personal, public, library]</td>\n","      <td>[(essential, 1), (every, 1), (personal, 1), (p...</td>\n","      <td>0.0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A2MVUWT453QH61</td>\n","      <td>7/7</td>\n","      <td>4.0</td>\n","      <td>Phlip Nel gives silly Seuss a serious treatment</td>\n","      <td>Theodore Seuss Geisel (1904-1991), aka &amp;quot;D...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[phlip, nel, gives, silly, seuss, serious, tre...</td>\n","      <td>[(phlip, 1), (nel, 1), (gives, 1), (silly, 1),...</td>\n","      <td>-1.0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>A22X4XUPKF66MR</td>\n","      <td>3/3</td>\n","      <td>4.0</td>\n","      <td>Good academic overview</td>\n","      <td>Philip Nel - Dr. Seuss: American IconThis is b...</td>\n","      <td>100.00</td>\n","      <td>positive</td>\n","      <td>[good, academic, overview]</td>\n","      <td>[(good, 1), (academic, 1), (overview, 1)]</td>\n","      <td>3.0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2e9ec9ff-f6bb-4297-9f5d-b217ddfc810e')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-2e9ec9ff-f6bb-4297-9f5d-b217ddfc810e button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-2e9ec9ff-f6bb-4297-9f5d-b217ddfc810e');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":74}]},{"cell_type":"code","source":["reviews = cleaned_reviews_df['review_text'].to_list()\n","\n","# Extract review summaries \n","summaries = cleaned_reviews_df['review_summary'].to_list()\n","\n","# Tokenize review summaries and store in a list (of lists). WARNING: Code takes about 5 minutes to run even for the summaries (unsure how much for reviews)\n","top_tokens_list = [tokenize_content(i) for i in summaries]\n","\n","# Flatten the list\n","top_tokens = [token for j in top_tokens_list for token in j]\n","\n","print(len(top_tokens))"],"metadata":{"id":"Icm6nUPXXoUy","executionInfo":{"status":"ok","timestamp":1671163871644,"user_tz":300,"elapsed":258339,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"32669756-fca9-4339-bcea-20a22d9f6d3a"},"execution_count":75,"outputs":[{"output_type":"stream","name":"stdout","text":["8998625\n"]}]},{"cell_type":"markdown","metadata":{"id":"uyxdJC6hey98"},"source":["###2.3 Word Clouds\n","Let's also just visualize some of these most common words as found in the tokens in a word cloud."]},{"cell_type":"code","source":["from wordcloud import WordCloud\n","from PIL import Image\n","# import matplotlib.pyplot as plt \n","word_dict = dict(Counter(top_tokens).most_common())\n","word_cloud = WordCloud(max_font_size=50, collocations = False, background_color = 'black').generate_from_frequencies(word_dict)\n","\n","plt.figure(figsize=(10,10))\n","plt.imshow(word_cloud, interpolation='bilinear')\n","plt.axis(\"off\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":310},"id":"cQj3aNxpLs5B","executionInfo":{"status":"ok","timestamp":1671164795296,"user_tz":300,"elapsed":2908,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"0c3f3589-9706-44ec-db53-300d204bacf1"},"execution_count":95,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 720x720 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["For a bit of fun, let's shape it in the letters of our favourite subject!"],"metadata":{"id":"CyapJ3hZLzTl"}},{"cell_type":"code","source":["from wordcloud import WordCloud\n","from PIL import Image\n","# import matplotlib.pyplot as plt \n","word_dict = dict(Counter(top_tokens).most_common())\n","mask = np.array(Image.open('/content/drive/MyDrive/CIS-logo.jpeg'))\n","word_cloud = WordCloud(max_font_size=50, collocations = False, background_color = 'black', mask=mask).generate_from_frequencies(word_dict)\n","\n","plt.figure(figsize=(10,10))\n","plt.imshow(word_cloud, interpolation='bilinear')\n","plt.axis(\"off\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":304},"id":"-rhUofbA54zn","executionInfo":{"status":"ok","timestamp":1671164768463,"user_tz":300,"elapsed":2596,"user":{"displayName":"Kofi Nightingale","userId":"03852095039738793853"}},"outputId":"14b2db8e-6fed-4f2e-cfdd-d00e956db90b"},"execution_count":94,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 720x720 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","metadata":{"id":"0TIiPwcDiHeC"},"source":["# Part 3: Data Modeling and Predictions\n","\n","What features can we use to predict the average rating of a book on Amazon?\n","\n","<br>\n","<center><img src = \"https://www.modernanalyst.com/Portals/0/Public%20Uploads/Fin387-Data-Mining-Visualization-Modeling.jpg\" width= \"400\" height='300' align =\"center\"/></center>\n","<br>\n","\n","Let's tackle this problem by trying both types of machine learning:\n","\n","\n","1.   Regression - quantitative response variable\n","2.   Classification - categorical response variable\n","\n","\n","\n"]},{"cell_type":"markdown","source":["### Structuring our Modeling Data Frame"],"metadata":{"id":"o2fybMYqOKo6"}},{"cell_type":"markdown","source":["Currently we still have two seperate dataframes. One dataframe `cleaned_books_df` is at the book-level, where each row represents a book. Whereas we still have a review-level dataframe `review_tokens_df`, where each row represents one review of a certain book.\n","\n","When we model we only want one row per book, so we will have to do group by's of our review-level dataframe, and then merge the two."],"metadata":{"id":"zF6EPQXYOdv0"}},{"cell_type":"markdown","source":["First, we will get book-level data from the `review_tokens_df` by averaging the numeric columns such as review_score (this will be our label for regression), `review_percent_helpful`, and `sentiment_score`. Extracting the number of reviews per book may also be helpful in prediction so that will also be extracted.\n","\n","Our label for regression will be `mean_review_score` but for classification we need to categorize this value."],"metadata":{"id":"Zflq959tS2OB"}},{"cell_type":"code","source":["review_avgs_df = review_tokens_df[[\"Title\", \"review_score\", \"review_percent_helpful\", \"sentiment_score\"]].groupby(by = \"Title\").mean().reset_index().rename(columns = {\"review_score\" : \"mean_review_score\", \"review_percent_helpful\" : \"mean_review_percent_helpful\", \"sentiment_score\" : \"mean_sentiment_score\"})"],"metadata":{"id":"fGfpx2USOQ03"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["review_count_df = review_tokens_df[[\"Title\", \"review_score\"]].groupby(by = \"Title\").count().reset_index().rename(columns = {\"review_score\" : \"review_count\"})"],"metadata":{"id":"RwnvT00GRvHJ"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["review_book_level_df = review_avgs_df.merge(review_count_df, on = 'Title')"],"metadata":{"id":"O8IQe_huS0n5"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["What is the distribution of mean review score?"],"metadata":{"id":"jAzW7HhoYIou"}},{"cell_type":"code","source":["percentiles = np.arange(0, 1.05, 0.05)\n","for i in range(len(percentiles)):\n","  print(round(percentiles[i]*100), \"th percentile:\", review_book_level_df.mean_review_score.quantile(percentiles[i]))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Xe8F5S-eYNAH","executionInfo":{"status":"ok","timestamp":1670885588110,"user_tz":300,"elapsed":190,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"2f59778a-02fe-4336-f067-43847380eecb"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["0 th percentile: 1.0\n","5 th percentile: 2.75\n","10 th percentile: 3.0\n","15 th percentile: 3.5\n","20 th percentile: 3.8\n","25 th percentile: 4.0\n","30 th percentile: 4.0\n","35 th percentile: 4.0\n","40 th percentile: 4.25\n","45 th percentile: 4.357142857142857\n","50 th percentile: 4.5\n","55 th percentile: 4.555555555555555\n","60 th percentile: 4.666666666666667\n","65 th percentile: 4.8\n","70 th percentile: 5.0\n","75 th percentile: 5.0\n","80 th percentile: 5.0\n","85 th percentile: 5.0\n","90 th percentile: 5.0\n","95 th percentile: 5.0\n","100 th percentile: 5.0\n"]}]},{"cell_type":"markdown","source":["It is important in machine learning to have somewhat balanced classes (ie. same number of observations in each class), however since this dataset has lots of books, we do not have to be as concerned having perfectly balanced classes. \n","\n","We notice from above that there are a lot of books (about 30% of them) that have a perfect 5.0 average rating. We are going to classify books as either \"Great\" or \"Not Great\". We will classify the top 50% of books as \"Great (1)\" (mean review score greater than 4.5) and \"Not Great (0)\" having a mean review score less than or equal to 4.5."],"metadata":{"id":"L7atIwNUa2GF"}},{"cell_type":"code","source":["review_book_level_df['mean_review_score_cat'] = review_book_level_df['mean_review_score'].apply(lambda x: \"1\" if x > 4.5 else \"0\")"],"metadata":{"id":"gqTB3N2ya1YN"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Although the classes are not perfectly balanced we think this method of ranking review scores will be the most representative of the titles."],"metadata":{"id":"QzK9fU4Xdj06"}},{"cell_type":"code","source":["review_book_level_df[[\"Title\", \"mean_review_score_cat\"]].groupby(\"mean_review_score_cat\").count()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":143},"id":"32uTa8Tgc6bw","executionInfo":{"status":"ok","timestamp":1670953873679,"user_tz":300,"elapsed":312,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"ad6ccafc-1073-4325-e3fb-5709030cdc20"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                        Title\n","mean_review_score_cat        \n","0                      114956\n","1                       97445"],"text/html":["\n","  <div id=\"df-7bb66eb1-e73b-46df-a7a8-15f23509188c\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","    </tr>\n","    <tr>\n","      <th>mean_review_score_cat</th>\n","      <th></th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>114956</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>97445</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7bb66eb1-e73b-46df-a7a8-15f23509188c')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-7bb66eb1-e73b-46df-a7a8-15f23509188c button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-7bb66eb1-e73b-46df-a7a8-15f23509188c');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":29}]},{"cell_type":"markdown","source":["Moving on to the `cleaned_books_df`, we definitely want to use the numeric features of `ratingsCount`, `num_authors`, and `num_categories` as predictors. As for the categorical features of `publisher`, `categories`, and `published_year` let's first see how many disctinct values we have, since in machine learning we have to one-hot encode our categorical features. "],"metadata":{"id":"1DQdzLTYzfKv"}},{"cell_type":"code","source":["n = len(pd.unique(cleaned_books_df[\"publisher\"]))\n","print(\"No.of.unique publisher values :\", n)\n","\n","n = len(pd.unique(cleaned_books_df[\"categories\"]))\n","print(\"No.of.unique categories values :\", n)\n","\n","n = len(pd.unique(cleaned_books_df[\"published_year\"]))\n","print(\"No.of.unique published_year values :\", n)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"eMD1vtlR0T5i","executionInfo":{"status":"ok","timestamp":1670953873880,"user_tz":300,"elapsed":203,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"02af7b88-d647-4dbe-983d-4fc194a24e0b"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["No.of.unique publisher values : 12223\n","No.of.unique categories values : 3482\n","No.of.unique published_year values : 166\n"]}]},{"cell_type":"markdown","source":["There are over 12,223 different publishers. But only 60 publishers have over 250 books published on Amazon and only 9 publishers have more than 1,000 books on Amazon. \n","\n","We do not want to one-hot encode 12,223 publishers, especially since the majority only publish a handfull of books, but we also do not want to lose some of this publisher information. We will one-hot the publishers with more than 250 books."],"metadata":{"id":"-GgaQ24w4Qs0"}},{"cell_type":"code","source":["pub_df = cleaned_books_df[[\"Title\", \"publisher\"]].groupby([\"publisher\"]).count().reset_index().sort_values(\"Title\", ascending=False)\n","pub_df.head(10)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"qv6wx07ViCS6","executionInfo":{"status":"ok","timestamp":1670953874135,"user_tz":300,"elapsed":257,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"b00d6a75-cb42-451c-d9f7-e6afbbc685a0"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                      publisher  Title\n","9971           SimonandSchuster   3425\n","8412                    Penguin   2799\n","4706              HarperCollins   2463\n","9425                  Routledge   2300\n","5691              JohnWileySons   1876\n","1945   CambridgeUniversityPress   1468\n","6599                  Macmillan   1209\n","2640         CourierCorporation   1113\n","8021              OpenRoadMedia    983\n","11426                   Vintage    955"],"text/html":["\n","  <div id=\"df-1d63f7c1-40b6-46fc-b8e0-b4fcfd07d6e1\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>publisher</th>\n","      <th>Title</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>9971</th>\n","      <td>SimonandSchuster</td>\n","      <td>3425</td>\n","    </tr>\n","    <tr>\n","      <th>8412</th>\n","      <td>Penguin</td>\n","      <td>2799</td>\n","    </tr>\n","    <tr>\n","      <th>4706</th>\n","      <td>HarperCollins</td>\n","      <td>2463</td>\n","    </tr>\n","    <tr>\n","      <th>9425</th>\n","      <td>Routledge</td>\n","      <td>2300</td>\n","    </tr>\n","    <tr>\n","      <th>5691</th>\n","      <td>JohnWileySons</td>\n","      <td>1876</td>\n","    </tr>\n","    <tr>\n","      <th>1945</th>\n","      <td>CambridgeUniversityPress</td>\n","      <td>1468</td>\n","    </tr>\n","    <tr>\n","      <th>6599</th>\n","      <td>Macmillan</td>\n","      <td>1209</td>\n","    </tr>\n","    <tr>\n","      <th>2640</th>\n","      <td>CourierCorporation</td>\n","      <td>1113</td>\n","    </tr>\n","    <tr>\n","      <th>8021</th>\n","      <td>OpenRoadMedia</td>\n","      <td>983</td>\n","    </tr>\n","    <tr>\n","      <th>11426</th>\n","      <td>Vintage</td>\n","      <td>955</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1d63f7c1-40b6-46fc-b8e0-b4fcfd07d6e1')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-1d63f7c1-40b6-46fc-b8e0-b4fcfd07d6e1 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-1d63f7c1-40b6-46fc-b8e0-b4fcfd07d6e1');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":31}]},{"cell_type":"code","source":["len(pub_df) - len(pub_df[pub_df['Title'] <= 250]) "],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"uBYUuuT8iG7C","executionInfo":{"status":"ok","timestamp":1670953874232,"user_tz":300,"elapsed":100,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"3ca05c60-4104-46a7-c3ae-5eeea3d8e4cc"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["60"]},"metadata":{},"execution_count":32}]},{"cell_type":"code","source":["#TODO: one-hot encode first 60 publishers\n","top_pub = pub_df[pub_df['Title'] > 250]['publisher'].tolist()\n","\n","for i in range(len(top_pub)):\n","  cleaned_books_df[top_pub[i]] = cleaned_books_df['publisher'].apply(lambda x: 1 if x == top_pub[i] else 0)"],"metadata":{"id":"2YEUSyi-YeEq"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Since `categories` also has a lot of distinct values (over 3,000), one-hot encoding every value would not be very smart here. We are going to do similar steps as before. There are 52 categories that have over 250 books in their category, and 54 books that have over 100 books in their category. Let's one-hot encode these top 54 categories."],"metadata":{"id":"80ye_Rfk8CYn"}},{"cell_type":"code","source":["cat_df = cleaned_books_df[[\"Title\", \"categories\"]].groupby([\"categories\"]).count().reset_index().sort_values(\"Title\", ascending=False)\n","cat_df.head(10)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"c17wbmuo8nHJ","executionInfo":{"status":"ok","timestamp":1670953883721,"user_tz":300,"elapsed":299,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"9d040bc4-462e-4550-93bf-0eacc427280f"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                         categories  Title\n","1948                    ['Fiction']  21096\n","2259                    ['History']   8136\n","3117                   ['Religion']   7919\n","2486           ['Juvenile Fiction']   6016\n","739   ['Biography & Autobiography']   5478\n","888        ['Business & Economics']   4942\n","1330                  ['Computers']   3852\n","3259             ['Social Science']   3353\n","2487        ['Juvenile Nonfiction']   2966\n","3181                    ['Science']   2292"],"text/html":["\n","  <div id=\"df-ed2be41c-c95b-4b1f-a2e3-cacfb3db8d80\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>categories</th>\n","      <th>Title</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>1948</th>\n","      <td>['Fiction']</td>\n","      <td>21096</td>\n","    </tr>\n","    <tr>\n","      <th>2259</th>\n","      <td>['History']</td>\n","      <td>8136</td>\n","    </tr>\n","    <tr>\n","      <th>3117</th>\n","      <td>['Religion']</td>\n","      <td>7919</td>\n","    </tr>\n","    <tr>\n","      <th>2486</th>\n","      <td>['Juvenile Fiction']</td>\n","      <td>6016</td>\n","    </tr>\n","    <tr>\n","      <th>739</th>\n","      <td>['Biography &amp; Autobiography']</td>\n","      <td>5478</td>\n","    </tr>\n","    <tr>\n","      <th>888</th>\n","      <td>['Business &amp; Economics']</td>\n","      <td>4942</td>\n","    </tr>\n","    <tr>\n","      <th>1330</th>\n","      <td>['Computers']</td>\n","      <td>3852</td>\n","    </tr>\n","    <tr>\n","      <th>3259</th>\n","      <td>['Social Science']</td>\n","      <td>3353</td>\n","    </tr>\n","    <tr>\n","      <th>2487</th>\n","      <td>['Juvenile Nonfiction']</td>\n","      <td>2966</td>\n","    </tr>\n","    <tr>\n","      <th>3181</th>\n","      <td>['Science']</td>\n","      <td>2292</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ed2be41c-c95b-4b1f-a2e3-cacfb3db8d80')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-ed2be41c-c95b-4b1f-a2e3-cacfb3db8d80 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-ed2be41c-c95b-4b1f-a2e3-cacfb3db8d80');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":34}]},{"cell_type":"code","source":["len(cat_df) - len(cat_df[cat_df['Title'] <= 100]) "],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"q6x9AnAglAhI","executionInfo":{"status":"ok","timestamp":1670953883723,"user_tz":300,"elapsed":6,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"ac22d235-25bf-4a09-f141-ef8c9e5dd66c"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["54"]},"metadata":{},"execution_count":35}]},{"cell_type":"code","source":["#TODO: one-hot encode first 54 categories\n","top_cat = cat_df[cat_df['Title'] > 100]['categories'].tolist()\n","\n","for i in range(len(top_cat)):\n","  cleaned_books_df[top_cat[i]] = cleaned_books_df['categories'].apply(lambda x: 1 if x == top_cat[i] else 0)"],"metadata":{"id":"31r2nWEalw4j"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Finally, looking at year there are significantly less distinct values as some of our other categorical variables (166), however, since we do not have an extrodinary amount of features, one-hot encoding each year and gaining 165 new features only relating to year will overwhelm the model and probably leave us with a lot of noise.\n","\n","The majority of books tended to be published between the late 1990s and early 2010s. There are 57 years with more than 100 books published. Let's one-hot encode these years."],"metadata":{"id":"kBuDig0E-pLa"}},{"cell_type":"code","source":["year_df = cleaned_books_df[[\"Title\", \"published_year\"]].groupby([\"published_year\"]).count().reset_index().sort_values(\"Title\", ascending=False)\n","year_df.head(10)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"no9lnmGl-o0D","executionInfo":{"status":"ok","timestamp":1670953892153,"user_tz":300,"elapsed":24,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"092abf76-8877-47eb-fac5-7c8dbab2e17a"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["    published_year  Title\n","145           2004   6295\n","146           2005   6159\n","144           2003   5857\n","143           2002   5447\n","141           2000   5119\n","142           2001   4960\n","153           2012   4926\n","154           2013   4654\n","138           1999   4528\n","152           2011   4312"],"text/html":["\n","  <div id=\"df-a0ffe184-3110-4d27-b01a-845cc92d98cd\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>published_year</th>\n","      <th>Title</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>145</th>\n","      <td>2004</td>\n","      <td>6295</td>\n","    </tr>\n","    <tr>\n","      <th>146</th>\n","      <td>2005</td>\n","      <td>6159</td>\n","    </tr>\n","    <tr>\n","      <th>144</th>\n","      <td>2003</td>\n","      <td>5857</td>\n","    </tr>\n","    <tr>\n","      <th>143</th>\n","      <td>2002</td>\n","      <td>5447</td>\n","    </tr>\n","    <tr>\n","      <th>141</th>\n","      <td>2000</td>\n","      <td>5119</td>\n","    </tr>\n","    <tr>\n","      <th>142</th>\n","      <td>2001</td>\n","      <td>4960</td>\n","    </tr>\n","    <tr>\n","      <th>153</th>\n","      <td>2012</td>\n","      <td>4926</td>\n","    </tr>\n","    <tr>\n","      <th>154</th>\n","      <td>2013</td>\n","      <td>4654</td>\n","    </tr>\n","    <tr>\n","      <th>138</th>\n","      <td>1999</td>\n","      <td>4528</td>\n","    </tr>\n","    <tr>\n","      <th>152</th>\n","      <td>2011</td>\n","      <td>4312</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a0ffe184-3110-4d27-b01a-845cc92d98cd')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-a0ffe184-3110-4d27-b01a-845cc92d98cd button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-a0ffe184-3110-4d27-b01a-845cc92d98cd');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":37}]},{"cell_type":"code","source":["len(year_df) - len(year_df[year_df['Title'] <= 100]) "],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"AijBFBG2mLIu","executionInfo":{"status":"ok","timestamp":1670953892154,"user_tz":300,"elapsed":23,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"44125041-07be-49f9-c1b6-0c25b6417446"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["57"]},"metadata":{},"execution_count":38}]},{"cell_type":"code","source":["#TODO: one-hot encode first 57 years\n","top_year = year_df[year_df['Title'] > 100]['published_year'].tolist()\n","\n","for i in range(len(top_year)):\n","  cleaned_books_df[top_year[i]] = cleaned_books_df['published_year'].apply(lambda x: 1 if x == top_year[i] else 0)"],"metadata":{"id":"RJRcSiNzmfFH"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Note: there is some ambiguity of what the feature `ratingsCount` actually represents, it seems like neither the number of reviews nor the average review score. We are curious to see if it does stand out within correlations, however our models will not contain this predictor as we have no interpretation of it."],"metadata":{"id":"2-hUkfzQgCIf"}},{"cell_type":"code","source":["books_df = cleaned_books_df.drop(columns = ['description', 'authors', 'publisher', 'categories', 'published_year'])\n","books_df.head(5)"],"metadata":{"id":"K1xjH06MfqjN","colab":{"base_uri":"https://localhost:8080/","height":543},"executionInfo":{"status":"ok","timestamp":1670953898768,"user_tz":300,"elapsed":23,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"1ffdf250-acdb-450d-e510-5cce98142c17"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title  ratingsCount  \\\n","0                           Dr. Seuss: American Icon           0.0   \n","1                      Whispers of the Wicked Saints           0.0   \n","2  The Church of Christ: A Biblical Ecclesiology ...           5.0   \n","3                           Saint Hyacinth of Poland           0.0   \n","4  Rising Sons and Daughters: Life Among Japan's ...           0.0   \n","\n","   num_authors  num_categories  SimonandSchuster  Penguin  HarperCollins  \\\n","0            1               1                 0        0              0   \n","1            1               1                 0        0              0   \n","2            1               1                 0        0              0   \n","3            1               1                 0        0              0   \n","4            1               1                 0        0              0   \n","\n","   Routledge  JohnWileySons  CambridgeUniversityPress  ...  1974  1973  1975  \\\n","0          0              0                         0  ...     0     0     0   \n","1          0              0                         0  ...     0     0     0   \n","2          0              0                         0  ...     0     0     0   \n","3          0              0                         0  ...     0     0     0   \n","4          0              0                         0  ...     0     0     0   \n","\n","   1972  1971  1969  1970  1967  1968  1965  \n","0     0     0     0     0     0     0     0  \n","1     0     0     0     0     0     0     0  \n","2     0     0     0     0     0     0     0  \n","3     0     0     0     0     0     0     0  \n","4     0     0     0     0     0     0     0  \n","\n","[5 rows x 175 columns]"],"text/html":["\n","  <div id=\"df-72f0f7f2-4de0-498a-8966-be0bda73ea07\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>ratingsCount</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","      <th>SimonandSchuster</th>\n","      <th>Penguin</th>\n","      <th>HarperCollins</th>\n","      <th>Routledge</th>\n","      <th>JohnWileySons</th>\n","      <th>CambridgeUniversityPress</th>\n","      <th>...</th>\n","      <th>1974</th>\n","      <th>1973</th>\n","      <th>1975</th>\n","      <th>1972</th>\n","      <th>1971</th>\n","      <th>1969</th>\n","      <th>1970</th>\n","      <th>1967</th>\n","      <th>1968</th>\n","      <th>1965</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Dr. Seuss: American Icon</td>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Whispers of the Wicked Saints</td>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>The Church of Christ: A Biblical Ecclesiology ...</td>\n","      <td>5.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>Saint Hyacinth of Poland</td>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>Rising Sons and Daughters: Life Among Japan's ...</td>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","<p>5 rows × 175 columns</p>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-72f0f7f2-4de0-498a-8966-be0bda73ea07')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-72f0f7f2-4de0-498a-8966-be0bda73ea07 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-72f0f7f2-4de0-498a-8966-be0bda73ea07');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":40}]},{"cell_type":"code","source":["review_book_level_df.head(5)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"Ck2MAGz9D48M","executionInfo":{"status":"ok","timestamp":1670953898769,"user_tz":300,"elapsed":22,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"3b2d7fc1-a526-4c0d-aec0-7d122eeef82b"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                                               Title  mean_review_score  \\\n","0           \" Film technique, \" and, \" Film acting \"              4.500   \n","1  \" We'll Always Have Paris\": The Definitive Gui...              5.000   \n","2  \"... And Poetry is Born ...\" Russian Classical...              4.000   \n","3       \"A Titanic hero\" Thomas Andrews, shipbuilder              4.875   \n","4  \"A Truthful Impression of the Country\": Britis...              4.000   \n","\n","   mean_review_percent_helpful  mean_sentiment_score  review_count  \\\n","0                        43.75                 0.000             2   \n","1                         0.00                 4.000             2   \n","2                         0.00                 4.000             1   \n","3                        75.00                 3.625             8   \n","4                       100.00                 0.000             1   \n","\n","  mean_review_score_cat  \n","0                     0  \n","1                     1  \n","2                     0  \n","3                     1  \n","4                     0  "],"text/html":["\n","  <div id=\"df-720ac4bc-9ef2-4d87-82ca-5d6d94260edf\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Title</th>\n","      <th>mean_review_score</th>\n","      <th>mean_review_percent_helpful</th>\n","      <th>mean_sentiment_score</th>\n","      <th>review_count</th>\n","      <th>mean_review_score_cat</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>\" Film technique, \" and, \" Film acting \"</td>\n","      <td>4.500</td>\n","      <td>43.75</td>\n","      <td>0.000</td>\n","      <td>2</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>\" We'll Always Have Paris\": The Definitive Gui...</td>\n","      <td>5.000</td>\n","      <td>0.00</td>\n","      <td>4.000</td>\n","      <td>2</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>\"... And Poetry is Born ...\" Russian Classical...</td>\n","      <td>4.000</td>\n","      <td>0.00</td>\n","      <td>4.000</td>\n","      <td>1</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>\"A Titanic hero\" Thomas Andrews, shipbuilder</td>\n","      <td>4.875</td>\n","      <td>75.00</td>\n","      <td>3.625</td>\n","      <td>8</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>\"A Truthful Impression of the Country\": Britis...</td>\n","      <td>4.000</td>\n","      <td>100.00</td>\n","      <td>0.000</td>\n","      <td>1</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-720ac4bc-9ef2-4d87-82ca-5d6d94260edf')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-720ac4bc-9ef2-4d87-82ca-5d6d94260edf button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-720ac4bc-9ef2-4d87-82ca-5d6d94260edf');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":41}]},{"cell_type":"markdown","source":["Now let's merge `review_book_level_df` and `books_df` on the book title to get our final dataframe for modeling."],"metadata":{"id":"hXNYnHhODFJH"}},{"cell_type":"code","source":["model_df = books_df.merge(review_book_level_df, on = 'Title', how = 'inner').drop(columns = 'Title')\n","model_df.head(5)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":299},"id":"U8waryAkDga0","executionInfo":{"status":"ok","timestamp":1670953899762,"user_tz":300,"elapsed":1013,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"bc24a6e7-13b0-485b-9ba2-d3c061ae1e57"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["   ratingsCount  num_authors  num_categories  SimonandSchuster  Penguin  \\\n","0           0.0            1               1                 0        0   \n","1           0.0            1               1                 0        0   \n","2           5.0            1               1                 0        0   \n","3           0.0            1               1                 0        0   \n","4           0.0            1               1                 0        0   \n","\n","   HarperCollins  Routledge  JohnWileySons  CambridgeUniversityPress  \\\n","0              0          0              0                         0   \n","1              0          0              0                         0   \n","2              0          0              0                         0   \n","3              0          0              0                         0   \n","4              0          0              0                         0   \n","\n","   Macmillan  ...  1969  1970  1967  1968  1965  mean_review_score  \\\n","0          0  ...     0     0     0     0     0           4.555556   \n","1          0  ...     0     0     0     0     0           3.718750   \n","2          0  ...     0     0     0     0     0           4.500000   \n","3          0  ...     0     0     0     0     0           4.500000   \n","4          0  ...     0     0     0     0     0           4.333333   \n","\n","   mean_review_percent_helpful  mean_sentiment_score  review_count  \\\n","0                    69.545556              2.000000             9   \n","1                    45.123750              0.593750            32   \n","2                    71.175000              0.750000             4   \n","3                   100.000000              2.000000             2   \n","4                    66.666667              2.333333             3   \n","\n","   mean_review_score_cat  \n","0                      1  \n","1                      0  \n","2                      0  \n","3                      0  \n","4                      0  \n","\n","[5 rows x 179 columns]"],"text/html":["\n","  <div id=\"df-9a4308ba-a01b-4da1-a59c-4acacf8d99cf\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>ratingsCount</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","      <th>SimonandSchuster</th>\n","      <th>Penguin</th>\n","      <th>HarperCollins</th>\n","      <th>Routledge</th>\n","      <th>JohnWileySons</th>\n","      <th>CambridgeUniversityPress</th>\n","      <th>Macmillan</th>\n","      <th>...</th>\n","      <th>1969</th>\n","      <th>1970</th>\n","      <th>1967</th>\n","      <th>1968</th>\n","      <th>1965</th>\n","      <th>mean_review_score</th>\n","      <th>mean_review_percent_helpful</th>\n","      <th>mean_sentiment_score</th>\n","      <th>review_count</th>\n","      <th>mean_review_score_cat</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>4.555556</td>\n","      <td>69.545556</td>\n","      <td>2.000000</td>\n","      <td>9</td>\n","      <td>1</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>3.718750</td>\n","      <td>45.123750</td>\n","      <td>0.593750</td>\n","      <td>32</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>5.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>4.500000</td>\n","      <td>71.175000</td>\n","      <td>0.750000</td>\n","      <td>4</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>4.500000</td>\n","      <td>100.000000</td>\n","      <td>2.000000</td>\n","      <td>2</td>\n","      <td>0</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>0.0</td>\n","      <td>1</td>\n","      <td>1</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>...</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>0</td>\n","      <td>4.333333</td>\n","      <td>66.666667</td>\n","      <td>2.333333</td>\n","      <td>3</td>\n","      <td>0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","<p>5 rows × 179 columns</p>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9a4308ba-a01b-4da1-a59c-4acacf8d99cf')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-9a4308ba-a01b-4da1-a59c-4acacf8d99cf button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-9a4308ba-a01b-4da1-a59c-4acacf8d99cf');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":42}]},{"cell_type":"code","source":["# for Grace so I don't have to run this code over and over again\n","# we can remove before submission\n","from google.colab import files\n","model_df.to_csv('model3.csv', encoding = 'utf-8-sig') \n","files.download('model3.csv')"],"metadata":{"id":"zXPCkBKMEU3i"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["model_df = pd.read_csv('/content/drive/MyDrive/model3.csv').iloc[: , 1:]"],"metadata":{"id":"7sru4IBfI09R"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["### Correlations"],"metadata":{"id":"9bFirB6J4qBf"}},{"cell_type":"markdown","source":["Let's first construct a correlation matrix of the features that were not one hot encoded."],"metadata":{"id":"he8Etzv-KlE9"}},{"cell_type":"code","source":["# Drop the labels\n","features = model_df[['ratingsCount', 'num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']]"],"metadata":{"id":"UwnEOGVz4tiV"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["corr_matrix = features.corr()\n","corr_matrix"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":237},"id":"JK8O2IKMOpdG","executionInfo":{"status":"ok","timestamp":1671016070183,"user_tz":300,"elapsed":214,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"af9b8b64-2411-4492-e3b7-a746f613277b"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["                             ratingsCount  num_authors  num_categories  \\\n","ratingsCount                     1.000000    -0.012981        0.004533   \n","num_authors                     -0.012981     1.000000       -0.008151   \n","num_categories                   0.004533    -0.008151        1.000000   \n","mean_review_percent_helpful     -0.025048    -0.002656        0.016698   \n","mean_sentiment_score            -0.002044    -0.000363       -0.003837   \n","review_count                     0.196716    -0.019734       -0.004007   \n","\n","                             mean_review_percent_helpful  \\\n","ratingsCount                                   -0.025048   \n","num_authors                                    -0.002656   \n","num_categories                                  0.016698   \n","mean_review_percent_helpful                     1.000000   \n","mean_sentiment_score                            0.030642   \n","review_count                                   -0.043217   \n","\n","                             mean_sentiment_score  review_count  \n","ratingsCount                            -0.002044      0.196716  \n","num_authors                             -0.000363     -0.019734  \n","num_categories                          -0.003837     -0.004007  \n","mean_review_percent_helpful              0.030642     -0.043217  \n","mean_sentiment_score                     1.000000     -0.002455  \n","review_count                            -0.002455      1.000000  "],"text/html":["\n","  <div id=\"df-f140929c-d592-454b-a0ac-c1c8e6a02ab3\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>ratingsCount</th>\n","      <th>num_authors</th>\n","      <th>num_categories</th>\n","      <th>mean_review_percent_helpful</th>\n","      <th>mean_sentiment_score</th>\n","      <th>review_count</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>ratingsCount</th>\n","      <td>1.000000</td>\n","      <td>-0.012981</td>\n","      <td>0.004533</td>\n","      <td>-0.025048</td>\n","      <td>-0.002044</td>\n","      <td>0.196716</td>\n","    </tr>\n","    <tr>\n","      <th>num_authors</th>\n","      <td>-0.012981</td>\n","      <td>1.000000</td>\n","      <td>-0.008151</td>\n","      <td>-0.002656</td>\n","      <td>-0.000363</td>\n","      <td>-0.019734</td>\n","    </tr>\n","    <tr>\n","      <th>num_categories</th>\n","      <td>0.004533</td>\n","      <td>-0.008151</td>\n","      <td>1.000000</td>\n","      <td>0.016698</td>\n","      <td>-0.003837</td>\n","      <td>-0.004007</td>\n","    </tr>\n","    <tr>\n","      <th>mean_review_percent_helpful</th>\n","      <td>-0.025048</td>\n","      <td>-0.002656</td>\n","      <td>0.016698</td>\n","      <td>1.000000</td>\n","      <td>0.030642</td>\n","      <td>-0.043217</td>\n","    </tr>\n","    <tr>\n","      <th>mean_sentiment_score</th>\n","      <td>-0.002044</td>\n","      <td>-0.000363</td>\n","      <td>-0.003837</td>\n","      <td>0.030642</td>\n","      <td>1.000000</td>\n","      <td>-0.002455</td>\n","    </tr>\n","    <tr>\n","      <th>review_count</th>\n","      <td>0.196716</td>\n","      <td>-0.019734</td>\n","      <td>-0.004007</td>\n","      <td>-0.043217</td>\n","      <td>-0.002455</td>\n","      <td>1.000000</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f140929c-d592-454b-a0ac-c1c8e6a02ab3')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-f140929c-d592-454b-a0ac-c1c8e6a02ab3 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-f140929c-d592-454b-a0ac-c1c8e6a02ab3');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":6}]},{"cell_type":"code","source":["# Set up the matplotlib figure\n","f, ax = plt.subplots(figsize=(10, 8))\n","\n","# Draw the heatmap with the mask and correct aspect ratio\n","sns.heatmap(corr_matrix, vmax=1, vmin = -1, center=0,\n","            square=True, linewidths=.5, cbar_kws={\"shrink\": .5}, cmap = 'RdBu')\n","plt.title(\"Correlation Heatmap for Amazon Book Data\")\n","plt.show()  "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":638},"id":"JyTYAqFMLQfk","executionInfo":{"status":"ok","timestamp":1671016070795,"user_tz":300,"elapsed":615,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"f707368e-a947-477a-bbfd-4b1fde3a5d7e"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 720x576 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["After reviewing the correlation matrix, we are not too concerned about mutlicolinearity (at least for the features that were not one-hot encoded) since none of the data have more extreme than 0.2 or -0.2. We do not need to do anything to address this issue initially, however the regularization we apply in the linear and logistic regression models will help with the mild multicolinearity. "],"metadata":{"id":"0oZ7X6kZPunU"}},{"cell_type":"markdown","source":["Now, let's look at the correlations between our regression response varaible `mean_review_score` and the features. When building linear regression, we do not want to overfit the data by putting in every feature. It is important to see the features that are correlated with the response variable so we can decide which features to include and which to remove."],"metadata":{"id":"4X_k0TM5Ktjc"}},{"cell_type":"code","source":["features_list = model_df.drop(columns = ['mean_review_score', 'mean_review_score_cat']).columns.tolist()"],"metadata":{"id":"IGR1g7qbQcT3"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["top_corr = []\n","for i in range(len(features_list)):\n","  if model_df['mean_review_score'].corr(model_df[features_list[i]]) > 0.005 or model_df['mean_review_score'].corr(model_df[features_list[i]]) < -0.005:\n","    #print(\"The correlation between mean_review_score and\", features_list[i], \"is\", round(model_df['mean_review_score'].corr(model_df[features_list[i]]), 2))\n","    top_corr.append(features_list[i])\n","\n","for i in range(len(features_list)):\n","  if model_df['mean_review_score'].corr(model_df[features_list[i]]) > 0.02 or model_df['mean_review_score'].corr(model_df[features_list[i]]) < -0.02:\n","    print(\"The correlation between mean_review_score and\", features_list[i], \"is\", round(model_df['mean_review_score'].corr(model_df[features_list[i]]), 2))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"q3Er73JdKtE3","executionInfo":{"status":"ok","timestamp":1671016073340,"user_tz":300,"elapsed":2547,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"8ba754d1-fe1b-415e-e14a-2a51d08b138d"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["The correlation between mean_review_score and JohnWileySons is -0.02\n","The correlation between mean_review_score and McGrawHillProfessional is -0.02\n","The correlation between mean_review_score and CengageLearning is -0.02\n","The correlation between mean_review_score and ['Fiction'] is -0.05\n","The correlation between mean_review_score and ['Religion'] is 0.06\n","The correlation between mean_review_score and ['Juvenile Fiction'] is 0.05\n","The correlation between mean_review_score and ['Computers'] is -0.08\n","The correlation between mean_review_score and ['Juvenile Nonfiction'] is 0.02\n","The correlation between mean_review_score and ['Cooking'] is 0.02\n","The correlation between mean_review_score and ['Technology & Engineering'] is -0.02\n","The correlation between mean_review_score and ['Self-Help'] is 0.02\n","The correlation between mean_review_score and ['Foreign Language Study'] is -0.03\n","The correlation between mean_review_score and ['Poetry'] is 0.03\n","The correlation between mean_review_score and ['Study Aids'] is -0.05\n","The correlation between mean_review_score and mean_review_percent_helpful is 0.09\n","The correlation between mean_review_score and mean_sentiment_score is 0.29\n"]}]},{"cell_type":"markdown","source":["Unfortunately for us, there are not very high correlations between our response variable and many of our predictors. The most extreme correlation is between mean review score and mean sentiment score. Hopefully we can still build effective models by implementing techniques we have learned in CIS 5450."],"metadata":{"id":"G3EZrqY2SNXt"}},{"cell_type":"markdown","source":["### Test Set"],"metadata":{"id":"GTkkH5WoTYS5"}},{"cell_type":"markdown","source":["Since we have so many books in our dataset, we are going to take out a true, final test set to run for our 3 final models. This data won't be used in either training the models or validating our hyperparameters. "],"metadata":{"id":"6BNVDIusTKut"}},{"cell_type":"code","source":["from sklearn.model_selection import train_test_split\n","features = model_df.drop(columns = ['mean_review_score', 'mean_review_score_cat', 'ratingsCount'])\n","labels = model_df[['mean_review_score', 'mean_review_score_cat']]\n","seed = 42\n","x_train, x_test_final, y_train, y_test_final = train_test_split(features, labels, random_state = seed, test_size = 0.1)"],"metadata":{"id":"09LSGkUMTD_w"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["### 3.1 Regression"],"metadata":{"id":"JmNDOvCgH4M1"}},{"cell_type":"markdown","source":["We are going to try to predict the average review score for a given book using linear regression. \n","\n","On top of just simple linear regression we are going to look at regularized linear regression (Lasso, Ridge, Elastic Net).\n","\n"],"metadata":{"id":"fkuKtf7t5Dlv"}},{"cell_type":"markdown","source":["#### 3.1.1 Linear Regression"],"metadata":{"id":"u8CjIG_E45kO"}},{"cell_type":"markdown","source":["First we will attempt a linear regression with all features."],"metadata":{"id":"0VlNr0ikNcNC"}},{"cell_type":"code","source":["#Get train and test splits\n","x_train_reg, x_test_reg, y_train_reg, y_test_reg = train_test_split(x_train, y_train[['mean_review_score']], random_state = seed, test_size = 0.2)"],"metadata":{"id":"PWAJBI5SVnw7"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# TO-DO: Import required libraries\n","from sklearn.linear_model import LinearRegression\n","\n","# TO-DO: Initialize model with default parameters and fit it on the training set\n","reg_full = LinearRegression()\n","reg_full.fit(x_train_reg, y_train_reg)\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_full = reg_full.score(x_test_reg, y_test_reg)\n","print(\"R-squared:\", round(score_full,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"mexLgisSN-Js","executionInfo":{"status":"ok","timestamp":1671016075813,"user_tz":300,"elapsed":2183,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"5917b6c4-b55b-4969-9e5a-89d340660080"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["R-squared: 0.1019\n"]}]},{"cell_type":"markdown","source":["This model does not perform very well. Let's try a subset of features that are the most correlated with `mean_review_score` to make sure we are not overfitting the data."],"metadata":{"id":"27RUhebsOwtz"}},{"cell_type":"code","source":["reg_corr = LinearRegression()\n","reg_corr.fit(x_train_reg[top_corr], y_train_reg)\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_corr = reg_corr.score(x_test_reg[top_corr], y_test_reg)\n","print(\"R-squared:\", round(score_corr,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"bHkCXa8eO4j_","executionInfo":{"status":"ok","timestamp":1671016077359,"user_tz":300,"elapsed":1549,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"6885068b-48c4-4b87-ac76-f3c07a924bae"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["R-squared: 0.1026\n"]}]},{"cell_type":"markdown","source":["There is only a marginal improvement in the $R^2$ of the result. Before moving on from classic linear regression, let's add some quadratic terms for the columns that are not one-hot encoded."],"metadata":{"id":"RbhDk_QaPkjx"}},{"cell_type":"code","source":["x_train_reg_pow = x_train_reg[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(2)\n","x_train_reg_pow.columns = ['num_authors_2', 'num_categories_2', 'mean_review_percent_helpful_2', 'mean_sentiment_score_2', 'review_count_2']\n","x_train_reg_pow = pd.concat([x_train_reg_pow, x_train_reg], axis=1)"],"metadata":{"id":"ostOYKZzRcva"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["x_test_reg_pow = x_test_reg[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(2)\n","x_test_reg_pow.columns = ['num_authors_2', 'num_categories_2', 'mean_review_percent_helpful_2', 'mean_sentiment_score_2', 'review_count_2']\n","x_test_reg_pow = pd.concat([x_test_reg_pow, x_test_reg], axis=1)"],"metadata":{"id":"MSU_zf6PS6Rm"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["reg_sq = LinearRegression()\n","reg_sq.fit(x_train_reg_pow, y_train_reg)\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_sq = reg_sq.score(x_test_reg_pow, y_test_reg)\n","print(\"R-squared:\", round(score_sq,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"WgkS-LK7Pj_h","executionInfo":{"status":"ok","timestamp":1671016079315,"user_tz":300,"elapsed":1671,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"68915628-6f51-4af0-b5a6-6567ec30a069"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["R-squared: 0.1227\n"]}]},{"cell_type":"markdown","source":["Since that was an improvement, let's try a degree 3 model."],"metadata":{"id":"SFusYcQzTMKB"}},{"cell_type":"code","source":["x_train_reg_pow3 = x_train_reg[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(3)\n","x_train_reg_pow3.columns = ['num_authors_3', 'num_categories_3', 'mean_review_percent_helpful_3', 'mean_sentiment_score_3', 'review_count_3']\n","x_train_reg_pow3 = pd.concat([x_train_reg_pow3, x_train_reg_pow], axis=1)\n","x_test_reg_pow3 = x_test_reg[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(3)\n","x_test_reg_pow3.columns = ['num_authors_3', 'num_categories_3', 'mean_review_percent_helpful_3', 'mean_sentiment_score_3', 'review_count_3']\n","x_test_reg_pow3 = pd.concat([x_test_reg_pow3, x_test_reg_pow], axis=1)"],"metadata":{"id":"ZsqO39r2TRn4"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["reg_cube = LinearRegression()\n","reg_cube.fit(x_train_reg_pow3, y_train_reg)\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_cube = reg_cube.score(x_test_reg_pow3, y_test_reg)\n","print(\"R-squared:\", round(score_cube,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"BdSbUSPuTjAx","executionInfo":{"status":"ok","timestamp":1671016081138,"user_tz":300,"elapsed":1598,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"08176649-c3ea-4c4e-f05c-135c7d702dcf"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["R-squared: 0.1225\n"]}]},{"cell_type":"markdown","source":["The model with the highest $R^2$ on the testing set had the features with correlations of 0.005 and greater or -0.005 and less, and all the non one-hot encoded features were squared.\n","\n","The $R^2$ tells us that 12.27% of the variability in mean review score for books from Amazon, can be explained by the linear relationship between mean review score and all the independent variables (features).\n","\n","Let's now look at regularized linear regression to see if we can improve our model."],"metadata":{"id":"E0e8UsrdN3QG"}},{"cell_type":"markdown","source":["####3.1.2 Regularized Linear Regression"],"metadata":{"id":"J6MPJS6K48vj"}},{"cell_type":"markdown","source":["We are going to now use regularized linear regression, which adds a penalty term to our model weights. There is Ridge Regression (L2), Lasso Regression (L1), and their trusty combination of Elastic Net Regression (L1 & L2)."],"metadata":{"id":"TVA8-gFNVux1"}},{"cell_type":"code","source":["from sklearn.linear_model import LinearRegression\n","from sklearn.linear_model import Lasso\n","from sklearn.linear_model import Ridge\n","from sklearn.linear_model import ElasticNet"],"metadata":{"id":"2OtiHQoPUnKn"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["####Lasso"],"metadata":{"id":"ZAOvAeOiSiyG"}},{"cell_type":"markdown","source":["First we are going to fit Lasso linear regression on our full dataset and report the testing $R^2$ for different values of the hyperparameter alpha. Alpha is a constant that determines the strength of the regularization, for example using alpha = 0.0 is just classic linear regression. "],"metadata":{"id":"svQLpPgCUx5o"}},{"cell_type":"code","source":["test_r_sq_full = []\n","alpha = [0.00001, 0.0001, 0.001, 0.01, 0.1, 1.0, 10.0]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  lasso_full = Lasso(alpha = alpha[i])\n","  lasso_full.fit(x_train_reg, y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_full = lasso_full.score(x_test_reg, y_test_reg)\n","  test_r_sq_full.append(score_full)\n","\n","test_r_sq_full"],"metadata":{"id":"QNy01lALLjHv","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1671016134229,"user_tz":300,"elapsed":13129,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"3caee039-9f78-4b89-c10d-0be0bf2ac695"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.10195563895335835,\n"," 0.10236892479212234,\n"," 0.0980769172057443,\n"," 0.08336919610734439,\n"," 0.07691731548784286,\n"," 0.004833346088102042,\n"," -2.604080225232508e-06]"]},"metadata":{},"execution_count":21}]},{"cell_type":"markdown","source":["We noticed that the highest $R^2$ falls around an alpha = 0.0001.\n","\n","Let's now focus in on alpha values around there. We will run these Lasso regressions on both the full dataset and the dataset reduced with the top correlated features."],"metadata":{"id":"OJxM7Vr1VXmm"}},{"cell_type":"code","source":["test_r_sq_full = []\n","alpha = [0.000009, 0.00001, 0.00002, 0.00003, 0.00004, 0.00005, 0.00006, 0.00007, 0.00008, 0.00009, 0.0001, 0.0002]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  lasso_full = Lasso(alpha = alpha[i])\n","  lasso_full.fit(x_train_reg, y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_full = lasso_full.score(x_test_reg, y_test_reg)\n","  test_r_sq_full.append(score_full)\n","\n","test_r_sq_full"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"RZRKJRtmSCaV","executionInfo":{"status":"ok","timestamp":1671025572285,"user_tz":300,"elapsed":42303,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"2936d227-4677-4735-9926-b54a844330f8"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.10194936309357461,\n"," 0.10195563895335835,\n"," 0.1019897268192167,\n"," 0.10200653172699947,\n"," 0.10206976898184128,\n"," 0.1021190071548751,\n"," 0.10217114272273553,\n"," 0.10223039264479372,\n"," 0.10228685012625971,\n"," 0.10233880979632248,\n"," 0.10236892479212234,\n"," 0.10236740244923015]"]},"metadata":{},"execution_count":99}]},{"cell_type":"code","source":["test_r_sq_red = []\n","alpha = [0.000009, 0.00001, 0.00002, 0.00003, 0.00004, 0.00005, 0.00006, 0.00007, 0.00008, 0.00009, 0.0001, 0.0002]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  lasso_red = Lasso(alpha = alpha[i])\n","  lasso_red.fit(x_train_reg[top_corr], y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_red = lasso_red.score(x_test_reg[top_corr], y_test_reg)\n","  test_r_sq_red.append(score_red)\n","\n","test_r_sq_red"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"VGDjdn4jSFuN","executionInfo":{"status":"ok","timestamp":1671025577499,"user_tz":300,"elapsed":5228,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"f97a88a5-3647-4ac4-d234-5cd8e541047e"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.1026067224820093,\n"," 0.10261048932729933,\n"," 0.10264332942297105,\n"," 0.10267225507103761,\n"," 0.1026950833039193,\n"," 0.10270647124249854,\n"," 0.10270877196399641,\n"," 0.10271034142922164,\n"," 0.10270676089339126,\n"," 0.10269621462556067,\n"," 0.10268467674043491,\n"," 0.1024304169454795]"]},"metadata":{},"execution_count":100}]},{"cell_type":"code","source":["lasso_df_full = pd.DataFrame({'R-Squared' : test_r_sq_full, 'alpha' : alpha})\n","lasso_df_full['Model'] = \"Full\"\n","lasso_df_red = pd.DataFrame({'R-Squared' : test_r_sq_red, 'alpha' : alpha})\n","lasso_df_red['Model'] = \"Reduced\"\n","lasso_df = pd.concat([lasso_df_full, lasso_df_red]).reset_index()"],"metadata":{"id":"R3YOKWMhSISt"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.lineplot(data = lasso_df, x = 'alpha', y = 'R-Squared', hue = 'Model')\n","plt.title(\"R-Squared of Lasso Regression at Given Alpha for Full and Reduced Model\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":421},"id":"WlESG-1kSO0A","executionInfo":{"status":"ok","timestamp":1671025577813,"user_tz":300,"elapsed":318,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"44c8b9ca-fe54-4fff-9198-343622c62a5f"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Text(0.5, 1.0, 'R-Squared of Lasso Regression at Given Alpha for Full and Reduced Model')"]},"metadata":{},"execution_count":102},{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Overall, we noticed that the reduced model tends to have higher $R^2$ than the full model for Lasso linear regression. We found the optimal alpha value to be 0.00007. This model has an $R^2$ of 0.10271."],"metadata":{"id":"dMCzfrCAVwrA"}},{"cell_type":"markdown","source":["#### Ridge"],"metadata":{"id":"Xg3Uxj4sSmHN"}},{"cell_type":"markdown","source":["Next we are going to fit Ridge linear regression on our full dataset and report the testing $R^2$ for different values of the hyperparameter alpha."],"metadata":{"id":"EHpCShmhW0XL"}},{"cell_type":"code","source":["test_r_sq_full = []\n","alpha = [0.00001, 0.0001, 0.001, 0.01, 0.1, 1.0, 10.0, 100.0, 1000.0]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  ridge_full = Ridge(alpha = alpha[i])\n","  ridge_full.fit(x_train_reg, y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_full = ridge_full.score(x_test_reg, y_test_reg)\n","  test_r_sq_full.append(score_full)\n","\n","test_r_sq_full"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"XAJSoAraSopa","executionInfo":{"status":"ok","timestamp":1671016191745,"user_tz":300,"elapsed":2957,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"32e4f343-2b13-490a-8ff5-8b63cbebfe15"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.10191153900812788,\n"," 0.1019115391608224,\n"," 0.10191154068796937,\n"," 0.10191155597952595,\n"," 0.1019117108906944,\n"," 0.1019134468681494,\n"," 0.1019405366179259,\n"," 0.10193515769673755,\n"," 0.09856111199494655]"]},"metadata":{},"execution_count":26}]},{"cell_type":"markdown","source":["We noticed that the highest $R^2$ falls somewhere between an alpha = 10.0 and alpha = 100.0.\n","\n","Let's now focus in on alpha values around there. We will run these Ridge regressions on both the full dataset and the dataset reduced with the top correlated features."],"metadata":{"id":"0J3ta2EpW-6_"}},{"cell_type":"code","source":["test_r_sq_full = []\n","alpha = [0,5,10,15,20,25,30,35,40,45,50,55,60,65,70,75,80,85,90]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  ridge_full = Ridge(alpha = alpha[i])\n","  ridge_full.fit(x_train_reg, y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_full = ridge_full.score(x_test_reg, y_test_reg)\n","  test_r_sq_full.append(score_full)\n","\n","test_r_sq_full"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"U_RJTKA6SrP6","executionInfo":{"status":"ok","timestamp":1671016199446,"user_tz":300,"elapsed":7715,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"34892ab4-b52b-4289-ad24-212846e899b4"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.10191153899116212,\n"," 0.1019241090370161,\n"," 0.1019405366179259,\n"," 0.10195714053076788,\n"," 0.10197227531953046,\n"," 0.10198522609184302,\n"," 0.10199572349865615,\n"," 0.10200371944294095,\n"," 0.10200927713047103,\n"," 0.10201251454997462,\n"," 0.10201357440563774,\n"," 0.10201260779174182,\n"," 0.10200976528681394,\n"," 0.10200519217395354,\n"," 0.10199902600463961,\n"," 0.10199139550996639,\n"," 0.10198242028834048,\n"," 0.10197221093575026,\n"," 0.10196086942053117]"]},"metadata":{},"execution_count":27}]},{"cell_type":"code","source":["test_r_sq_red = []\n","alpha = [0,5,10,15,20,25,30,35,40,45,50,55,60,65,70,75,80,85,90]\n","\n","for i in range(len(alpha)):\n","  # TO-DO: Initialize model with default parameters and fit it on the training set\n","  ridge_red = Ridge(alpha = alpha[i])\n","  ridge_red.fit(x_train_reg[top_corr], y_train_reg)\n","\n","  # TO-DO: Find the R-squared score and store the value\n","  score_red = ridge_red.score(x_test_reg[top_corr], y_test_reg)\n","  test_r_sq_red.append(score_red)\n","\n","test_r_sq_red"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"re78tuADStRT","executionInfo":{"status":"ok","timestamp":1671016204971,"user_tz":300,"elapsed":5528,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"fa94ea13-fd88-4e1c-8a82-296543dc8862"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.10256893958561719,\n"," 0.10259062747677028,\n"," 0.10260626644600734,\n"," 0.10261670917353871,\n"," 0.10262265743062227,\n"," 0.10262469580410472,\n"," 0.10262331633518407,\n"," 0.10261893689454449,\n"," 0.10261191513768031,\n"," 0.10260255927613882,\n"," 0.10259113651190865,\n"," 0.10257787972783816,\n"," 0.10256299285669634,\n"," 0.10254665523524464,\n"," 0.10252902516882545,\n"," 0.10251024287482613,\n"," 0.10249043293232507,\n"," 0.10246970633535468,\n"," 0.10244816222516417]"]},"metadata":{},"execution_count":28}]},{"cell_type":"code","source":["ridge_df_full = pd.DataFrame({'R-Squared' : test_r_sq_full, 'alpha' : alpha})\n","ridge_df_full['Model'] = \"Full\"\n","ridge_df_red = pd.DataFrame({'R-Squared' : test_r_sq_red, 'alpha' : alpha})\n","ridge_df_red['Model'] = \"Reduced\"\n","ridge_df = pd.concat([ridge_df_full, ridge_df_red]).reset_index()"],"metadata":{"id":"3AOW1IddSvJd"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.lineplot(data = ridge_df, x = 'alpha', y = 'R-Squared', hue = 'Model')\n","plt.title(\"R-Squared of Ridge Regression at Given Alpha for Full and Reduced Model\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":421},"id":"IzTf9eIASxDS","executionInfo":{"status":"ok","timestamp":1671016205376,"user_tz":300,"elapsed":419,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"95ec0212-372f-4cae-a7be-30e840bb3337"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["Text(0.5, 1.0, 'R-Squared of Ridge Regression at Given Alpha for Full and Reduced Model')"]},"metadata":{},"execution_count":30},{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Overall, we noticed that the reduced model tends to have higher $R^2$ than the full model for Ridge linear regression, just like with Lasso. We found the optimal alpha value to be 25. This model has an $R^2$ of 0.10262, which is slightly lower than with Lasso regression."],"metadata":{"id":"xlWEG2TDXWOh"}},{"cell_type":"markdown","source":["#### Elastic Net"],"metadata":{"id":"ZEb_l5uTSyyM"}},{"cell_type":"markdown","source":["Our final step for linear regression is looking at the elactic net model which combines both Lasso and Ridge regression penalties. We now have 2 hyperparameters to tune, alpha like before, and now also the l1_ratio. When the l1_ration = 0, we have Ridge regression and when l1_ratio = 1 we have Lasso regression."],"metadata":{"id":"ieiynQ24XrSr"}},{"cell_type":"code","source":["alpha = [0.00001, 0.0001, 0.001, 0.01, 0.1, 1.0, 10.0]\n","l1 = [0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]\n","test_r_sq_red = []\n","\n","for i in range(len(alpha)):\n","  for j in range(len(l1)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    en_red = ElasticNet(alpha = alpha[i], l1_ratio = l1[j])\n","    en_red.fit(x_train_reg[top_corr], y_train_reg)\n","\n","    # TO-DO: Find the R-squared score and store the value\n","    score_red = en_red.score(x_test_reg[top_corr], y_test_reg)\n","    test_r_sq_red.append(score_red)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"oSLrtnOnS2o5","executionInfo":{"status":"ok","timestamp":1671016474000,"user_tz":300,"elapsed":268628,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"c97bcce5-4200-45df-e2a4-a6b714a28d09"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stderr","text":["/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.393e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.393e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.399e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.419e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.457e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.541e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_coordinate_descent.py:647: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations, check the scale of the features or consider increasing regularisation. Duality gap: 2.664e+04, tolerance: 5.451e+00 Linear regression models with null weight for the l1 regularization term are more efficiently fitted using one of the solvers implemented in sklearn.linear_model.Ridge/RidgeCV instead.\n","  model = cd_fast.enet_coordinate_descent(\n"]}]},{"cell_type":"code","source":["rsq = np.array(test_r_sq_red)\n","rsq = rsq.reshape(7,11)\n","elastic_net_df = pd.DataFrame(rsq, index = alpha, columns = l1)"],"metadata":{"id":"5dqMkSycTGzi"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["elastic_net_df"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":269},"id":"LLxq6tBoTLw0","executionInfo":{"status":"ok","timestamp":1671016474001,"user_tz":300,"elapsed":13,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"4cc04f3f-e8ea-4b01-e4b4-8824d9fc8af4"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["               0.0       0.1       0.2       0.3       0.4       0.5  \\\n","0.00001   0.102573  0.102577  0.102581  0.102585  0.102589  0.102593   \n","0.00010   0.102603  0.102633  0.102657  0.102679  0.102697  0.102705   \n","0.00100   0.102461  0.102268  0.101909  0.101514  0.101068  0.100492   \n","0.01000   0.099048  0.095216  0.092745  0.091722  0.090967  0.090042   \n","0.10000   0.090048  0.082443  0.081503  0.081367  0.081113  0.080738   \n","1.00000   0.076678  0.063167  0.043473  0.014235  0.005272  0.005281   \n","10.00000  0.029282  0.004812  0.001379 -0.000003 -0.000003 -0.000003   \n","\n","               0.6       0.7       0.8       0.9       1.0  \n","0.00001   0.102596  0.102600  0.102604  0.102607  0.102610  \n","0.00010   0.102706  0.102706  0.102703  0.102694  0.102685  \n","0.00100   0.099909  0.099319  0.098846  0.098404  0.098055  \n","0.01000   0.088928  0.087605  0.086322  0.084953  0.083369  \n","0.10000   0.080239  0.079611  0.078851  0.077954  0.076917  \n","1.00000   0.005259  0.005203  0.005114  0.004991  0.004833  \n","10.00000 -0.000003 -0.000003 -0.000003 -0.000003 -0.000003  "],"text/html":["\n","  <div id=\"df-63e1adb1-70b8-4be0-9efa-6879bb3e26c3\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>0.0</th>\n","      <th>0.1</th>\n","      <th>0.2</th>\n","      <th>0.3</th>\n","      <th>0.4</th>\n","      <th>0.5</th>\n","      <th>0.6</th>\n","      <th>0.7</th>\n","      <th>0.8</th>\n","      <th>0.9</th>\n","      <th>1.0</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0.00001</th>\n","      <td>0.102573</td>\n","      <td>0.102577</td>\n","      <td>0.102581</td>\n","      <td>0.102585</td>\n","      <td>0.102589</td>\n","      <td>0.102593</td>\n","      <td>0.102596</td>\n","      <td>0.102600</td>\n","      <td>0.102604</td>\n","      <td>0.102607</td>\n","      <td>0.102610</td>\n","    </tr>\n","    <tr>\n","      <th>0.00010</th>\n","      <td>0.102603</td>\n","      <td>0.102633</td>\n","      <td>0.102657</td>\n","      <td>0.102679</td>\n","      <td>0.102697</td>\n","      <td>0.102705</td>\n","      <td>0.102706</td>\n","      <td>0.102706</td>\n","      <td>0.102703</td>\n","      <td>0.102694</td>\n","      <td>0.102685</td>\n","    </tr>\n","    <tr>\n","      <th>0.00100</th>\n","      <td>0.102461</td>\n","      <td>0.102268</td>\n","      <td>0.101909</td>\n","      <td>0.101514</td>\n","      <td>0.101068</td>\n","      <td>0.100492</td>\n","      <td>0.099909</td>\n","      <td>0.099319</td>\n","      <td>0.098846</td>\n","      <td>0.098404</td>\n","      <td>0.098055</td>\n","    </tr>\n","    <tr>\n","      <th>0.01000</th>\n","      <td>0.099048</td>\n","      <td>0.095216</td>\n","      <td>0.092745</td>\n","      <td>0.091722</td>\n","      <td>0.090967</td>\n","      <td>0.090042</td>\n","      <td>0.088928</td>\n","      <td>0.087605</td>\n","      <td>0.086322</td>\n","      <td>0.084953</td>\n","      <td>0.083369</td>\n","    </tr>\n","    <tr>\n","      <th>0.10000</th>\n","      <td>0.090048</td>\n","      <td>0.082443</td>\n","      <td>0.081503</td>\n","      <td>0.081367</td>\n","      <td>0.081113</td>\n","      <td>0.080738</td>\n","      <td>0.080239</td>\n","      <td>0.079611</td>\n","      <td>0.078851</td>\n","      <td>0.077954</td>\n","      <td>0.076917</td>\n","    </tr>\n","    <tr>\n","      <th>1.00000</th>\n","      <td>0.076678</td>\n","      <td>0.063167</td>\n","      <td>0.043473</td>\n","      <td>0.014235</td>\n","      <td>0.005272</td>\n","      <td>0.005281</td>\n","      <td>0.005259</td>\n","      <td>0.005203</td>\n","      <td>0.005114</td>\n","      <td>0.004991</td>\n","      <td>0.004833</td>\n","    </tr>\n","    <tr>\n","      <th>10.00000</th>\n","      <td>0.029282</td>\n","      <td>0.004812</td>\n","      <td>0.001379</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","      <td>-0.000003</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-63e1adb1-70b8-4be0-9efa-6879bb3e26c3')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-63e1adb1-70b8-4be0-9efa-6879bb3e26c3 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-63e1adb1-70b8-4be0-9efa-6879bb3e26c3');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":33}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(elastic_net_df, cmap = 'RdBu')\n","plt.title(\"Heatmap for R^2 Given Elastic Net Parameters\")\n","plt.xlabel(\"L1 Ratio\")\n","plt.ylabel(\"Alpha\")\n","plt.show() "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"ncZ8gdBNTKl8","executionInfo":{"status":"ok","timestamp":1671016474395,"user_tz":300,"elapsed":262,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"ec002bb8-2d7c-4af2-fa5e-be626217d385"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["The highest $R^2$ for Elastic Net is 0.10271 for an alpha = 0.0001 and an L1 ratio = 0.7 or 0.6. This is identical to our highest $R^2$ for Lasso regression, and higher than that of the Ridge regression.\n","\n","We will conclude that our best regularized regression model is Lasso (rather than Elastic Net since it's simpler) with an alpha = 0.00007. For our final regression model, we will fit this Lasso model on our quadratic data."],"metadata":{"id":"dr8iGR5gYumG"}},{"cell_type":"code","source":["lasso_sq = Lasso(alpha = 0.00007)\n","lasso_sq.fit(x_train_reg_pow, y_train_reg)\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_sq = lasso_sq.score(x_test_reg_pow, y_test_reg)\n","print(\"R-squared:\", round(score_sq,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"YiZJrAmlZ7pi","executionInfo":{"status":"ok","timestamp":1671018200684,"user_tz":300,"elapsed":8347,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"d02279ae-8dc1-44de-d4cd-7aa5728fbdde"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["R-squared: 0.123\n"]}]},{"cell_type":"markdown","source":["Coming in with an $R^2$ of 0.123, this is the best regression model!\n","\n","Overall, regression does not seem to do very well with the Amazon books data. It is very hard to predict the mean review score for a book with any or all of the features we have. We believe that this partially has to do with the sheer quantity of books, there might be more noise than signal in our dataset.\n","\n","We are excited to begin looking into classification models, as predicting a binary output relating to mean review score may have more success than predicting the mean review score numerically."],"metadata":{"id":"t6Jd_NcNaNOD"}},{"cell_type":"markdown","source":["###3.2 Classification"],"metadata":{"id":"aV8lhCHzLjxM"}},{"cell_type":"markdown","source":["When we run classification models we have a categorical label. We have labeled a book with a mean review score greater than 4.5 as a 1, and all other books as a 0. There is a 50/50 split of observations in each of the two classes. We will look at both Logistic Regression (with and without regularization), as well as Random Forrests to find an optimal classification model.\n","\n","*Note:* there are many metrics you can use to evaluate the performance of classification models (percision, recall, F1, auc, etc.), however since our classes are perfectly balanced and we don't have any strong importance placed on either of our labels, we chose accuracy to be the metric of comparison."],"metadata":{"id":"jMBL0gwca-Cr"}},{"cell_type":"markdown","source":["####3.2.1 Logistic Regression"],"metadata":{"id":"N52X5hVr4va0"}},{"cell_type":"code","source":["x_train_class, x_test_class, y_train_class, y_test_class = train_test_split(x_train, y_train[['mean_review_score_cat']], random_state = seed, test_size = 0.2)"],"metadata":{"id":"MEQSnPkoLmbu"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Let's first train and test Logistic Regression on both our full and reduced datasets and compare the test accuracy."],"metadata":{"id":"4lI9UmPvbdv8"}},{"cell_type":"code","source":["# TO-DO: Import required libraries\n","from sklearn.linear_model import LogisticRegression\n","\n","# TO-DO: Initialize model with default parameters and fit it on the training set\n","logreg = LogisticRegression(max_iter=1000, penalty='none')\n","logreg.fit(x_train_class, y_train_class.values.ravel())\n","\n","# TO-DO: Find the accuracy and store the value in `log_acc`\n","log_acc = logreg.score(x_test_class, y_test_class.values.ravel())\n","log_acc"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"OIRV--x8Td0V","executionInfo":{"status":"ok","timestamp":1671016506464,"user_tz":300,"elapsed":31909,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"71fc0f7d-2277-4522-e06c-64e8abcfaf53"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.6263099834528406"]},"metadata":{},"execution_count":36}]},{"cell_type":"code","source":["# TO-DO: Initialize model with default parameters and fit it on the training set\n","logreg = LogisticRegression(max_iter=1000, penalty='none')\n","logreg.fit(x_train_class[top_corr], y_train_class.values.ravel())\n","\n","# TO-DO: Find the accuracy and store the value in `log_acc`\n","log_acc_red = logreg.score(x_test_class[top_corr], y_test_class.values.ravel())\n","log_acc_red"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"nQSHaa9GTfb_","executionInfo":{"status":"ok","timestamp":1671016522468,"user_tz":300,"elapsed":16019,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"824c3ad1-cd96-46c0-eb25-5ca7b1cba928"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["0.6253906968192683"]},"metadata":{},"execution_count":37}]},{"cell_type":"markdown","source":["It looks like the full model tends to perfrom better than the reduced model, which is a difference from our linear regression. From now on let's only look at the regressions for the full dataset.\n","\n","Next, we want to compare Ridge and Lasso regularization. We are going to need to hyperparameter tune 'C', which is the inverse of the regularization strength. Think of it as 1/alpha when comparing this case to our regularized linear regression models."],"metadata":{"id":"4_wxeJvyd5kL"}},{"cell_type":"code","source":["#Ridge\n","C = [0.0001, 0.001, 0.01, 0.1, 1.0, 10.0, 100.0] #inverse of regularization strength\n","test_acc = []\n","\n","for i in range(len(C)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    logreg = LogisticRegression(max_iter=1000, penalty='l2', C = C[i])\n","    logreg.fit(x_train_class, y_train_class.values.ravel())\n","\n","    # TO-DO: Find the  accuracy score and store the value\n","    score = logreg.score(x_test_class, y_test_class.values.ravel())\n","    test_acc.append(score)\n","\n","test_acc"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"oBkKgoYgThBc","executionInfo":{"status":"ok","timestamp":1671016679406,"user_tz":300,"elapsed":156953,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"ac961a1e-6c57-46f5-f1cf-09e5b050da06"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.6138076852362566,\n"," 0.6282864497150211,\n"," 0.629021879021879,\n"," 0.6267236624379482,\n"," 0.6268155911013054,\n"," 0.6264938407795551,\n"," 0.6267236624379482]"]},"metadata":{},"execution_count":38}]},{"cell_type":"code","source":["#Lasso\n","C = [0.0001, 0.001, 0.01, 0.1, 1.0, 10.0, 100.0] #inverse of regularization strength\n","test_acc_lasso = []\n","\n","for i in range(len(C)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    logreg = LogisticRegression(max_iter=1000, penalty='l1', C = C[i], solver = 'liblinear')\n","    logreg.fit(x_train_class, y_train_class.values.ravel())\n","\n","    # TO-DO: Find the  accuracy score and store the value\n","    score = logreg.score(x_test_class, y_test_class.values.ravel())\n","    test_acc_lasso.append(score)\n","\n","test_acc_lasso"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"3IOlFpuKTlKh","executionInfo":{"status":"ok","timestamp":1671016691996,"user_tz":300,"elapsed":12593,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"2c5924c8-e876-4c9a-97d0-2e15126f420a"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["[0.5649476006618864,\n"," 0.6204725133296561,\n"," 0.6277808420665564,\n"," 0.6280566280566281,\n"," 0.6266776981062695,\n"," 0.6272752344180915,\n"," 0.6272752344180915]"]},"metadata":{},"execution_count":39}]},{"cell_type":"code","source":["ridge_df = pd.DataFrame({'Accuracy' : test_acc, 'C' : C})\n","ridge_df['Regularization'] = \"Ridge\"\n","lasso_df = pd.DataFrame({'Accuracy' : test_acc_lasso, 'C' : C})\n","lasso_df['Regularization'] = \"Lasso\"\n","df = pd.concat([ridge_df, lasso_df]).reset_index()"],"metadata":{"id":"u3t2JKrgTpXn"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.barplot(data = df, x = 'C', y = 'Accuracy', hue = 'Regularization')\n","plt.title(\"Accuracy of Logistic Regression Lasso vs. Ridge\")\n","plt.legend(loc='lower right')\n","plt.ylim((0.55,0.635))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":421},"id":"cWomtAENTrJ9","executionInfo":{"status":"ok","timestamp":1671018815449,"user_tz":300,"elapsed":872,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"c755fcde-cd4f-4add-e90a-7c8cf5d9ca66"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(0.55, 0.635)"]},"metadata":{},"execution_count":51},{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAfgAAAGDCAYAAADHzQJ9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3debgedX3//+fLBAgIyhYRCBiqoFhBhIhbUQSRuIGKRXABrIC2RVurKNR+aaXutNX2J9aioqIiKFYaWhQ3FKuiCRZZgmhEhLBIDCCyBAi8f3/MHLg5npPcWeacZPJ8XNd95Z79/Zl7cl5nPjP3nFQVkiSpXx422QVIkqTVz4CXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4aZQkz0ryiyS3J3npBG3zY0n+30ost31b55Qu6lpTJflqksMnu44+S7JXkiuXMf3TSd49kTVpxRjw+gNJvpPkliQbTHYtk+RE4CNVtXFVnT16YpKrkzxvdW6wqt5YVf+4vPlGb7uqrmnrvG9FtpfkiCT3tb8c3Jbkp0levDK1T4aqekFVfWZ1r3ddCq0k/5Dk3vYYuDXJD5I8Y2R6VX2vqh4/mTVq1RjweogkM4G9gAIOmOBtT53I7S3DY4DLJ7uICfDDqtoY2BT4KHBGkk1X90bWtd6FtcyZ7TGwJXA+8KVJrkerkQGv0Q4DLgQ+DTykCzTJdkn+M8miJIuTfGRg2lFJrkjy+yTzk+zejq8kjxuY74EzpCR7J1mY5B1JbgQ+lWSzJP/dbuOW9v2MgeU3T/KpJNe3089ux1+W5CUD862X5LdJnjJWI9t6FyS5OcmcJNu0438J/BFwTntmM3QvRpINkny4re369v0GA9PfnuSGdtqRg/tm1H7Zsm33rW1930vysCSfBbYfqO3tSWa265m6rP2zLFV1P/BZ4OHAjgNt+ack1yT5TXsJYcMVaMu/Jzk3yR3Ac5Nsk+TL7ef6qyRvHljXnknmtT0Jv0nyL+34aUk+1x5rtyaZm2Srdtp3khzZvn9Ykr9L8uskNyU5Lckj22kj++fwti2/TfLOYT/TUZ/vvya5tq3zoiR7rWIbtmmPvZvbY/Gocbb7tCQ3ZuAXpSQvS3LJsra9IqpqKfB5YNsk09v17p1k4cA2n5LkJ2n+j58JTBtV57KOiWUeT+qGAa/RDqP5j/55YP+BH0ZTgP8Gfg3MBLYFzmin/SnwD+2yj6A581885PYeDWxOc9Z8NM0x+al2eHvgLuAjA/N/FtgI+GPgUcCH2vGnAa8ZmO+FwA1V9X+jN5hkH+B9wMHA1m2bzgCoqscC1wAvabu+7x6yHQDvBJ4O7AY8GdgT+Lt2m7OBvwGeBzwO2HsZ63krsBCYDmwF/G1TWr12VG0fHGPZ8fbPuNrP9nXAvTT7AuD9wE5tWx5H83mfsAJteRXwHmAT4AfAOcBP2/XsC/x1kv3bef8V+NeqegTwWOCL7fjDgUcC2wFbAG+kOR5GO6J9PZfml7ONeegxA/AnwOPbbZ+QZOdl7JLxzKXZH5sDpwNfSjIScivThjNoPudtgFcA722PzYeoqh8BdwCD017V1rCsbQ8tyfo0/38XA7eMM/1smuNrc5oz/YMGpi/vmBj3eFKHqsqXL6oKmh+C9wJbtsM/A97Svn8GsAiYOsZy5wF/Nc46C3jcwPCngXe37/cG7gGmLaOm3YBb2vdbA/cDm40x3zbA74FHtMNnAW8fZ52fBD44MLxx2+6Z7fDVwPOWUdOY04FfAi8cGN4fuLp9fyrwvoFpjxvcN6P2y4nAfw3ut/G2TfPLVgFTl7V/xljPEcBS4Na27XcBB7fTQhMojx2Y/xnAr1agLacNTH8acM2o7R8PfKp9fwHwrpHjbmCeP6P55WDXMer/DnBk+/5bwF8MTHt826apA/tnxsD0HwOHjLNfHvgchtiHtwBPXpk20AT+fcAmA+PeB3x6nG29Gzi1fb9J+/k8ZlnbHqL+f6D5/3drW8tiYO+B6XsDC9v3zwauBzIw/QcDx+y4x8Tyjidf3b08g9egw4GvV9Vv2+HTebCbfjvg19V05Y22HU24rYxFVbVkZCDJRkn+o+1uvY3mh9em7VnmdsDNVfUHZxhVdT3wfeCgNNeRX0DTCzGWbXjwTJWqup3mh9u2K9mGMdfbvt9mYNq1A9MG3492ErAA+HqSq5IcN+T2x90/47iwqjYFNgPm0Nx7AU3PwUbARW238q3A19rxMFxbBsc9BthmZF3t+v6WpncC4PU0Z3c/a7uwR272+yzNL49ntN2+H0yy3hjbGmu/Tx1YP8CNA+/vpPmlboUkeVuay1C/a9vwSJpr1yvThm1oPqvfj6p7vGPwdODlaS75vBz4SVWNtHm8bQ/ji+0xsBVwGbDHOPNtA1xXbToP1Ds4fbxjYnnHkzqyptzUpEnWXg87GJiS5no4wAY04fpkmv+w2yeZOkbIX0vTNTiWO2n+c494NE235IjRf87wrTRnYE+rqhuT7Ab8H81ZwLXA5kk2rapbx9jWZ4AjaY7rH1bVdePUdD1N6ACQ5OE03afjzT+skfWO3KC3fTsO4AZgxsC82423kvaH/luBtyZ5EvDtJHOr6lv84f4atLz9M972bk/y58BVSU6l6Uq/C/jjcfbhMG0ZrPNamrO1HcfZ/i+AQ5M8jCa8zkqyRVXdQXNm+q40N3+eC1xJ0wMz6CGfJ81+Xwr8ZlSdK6293v52mi7+y6vq/iS30ByXK9OGr9N8VpsMhPz2jHMMVtX8JL+m+cV1sHt+edseSlX9NsnRwLwkp1fVDaNmuYHm+nwGQn57HvzFflnHxG9Z9vGkjngGrxEvpemmeyJNt/huwM7A92iuzf2Y5j/x+5M8vL156Fntsp8A3pZkjzQel2TkB+7FwKuSTGmv0z1nOXVsQvPD4NYkmwN/PzKh/aHzVeCjaW7GWy/JsweWPRvYHfgrmmvy4/kC8Loku7VnRO8FflRVVy+ntkHrtftg5DW1Xe/fJZmeZEuaa4yfa+f/YrvNnZNsBIz7nfckL273YYDf0Xwu97eTf0NznfkPDLF/xlVVN9N8jidUc9Pdx4EPJXlUW9O2A9fMh25L68fA79PcTLlheyw8KclT23W/Jsn0drsjv5jcn+S5SXZpe29uo+l2v3+M9X8BeEuSHZJsTPN5njlOb9Mwpoz6bNenOS6X0l6mSnICzf0mrEwbqupami7u97Xb2JXmTPxzjO90mmP72Qzc7T7etle00VV1JU1vw9vHmPzDtv1vbo+rl9PcYzJi3GNiiONJHTHgNeJwmmui11TVjSMvmpuVXk1zpvISmmtq19Cchb8SoKq+RHND1ek018HPprkRB5ofSC+h+cHz6nbasnwY2JDmt/4LabryBr2W5ofkz4CbgL8emVBVdwFfBnYA/nO8DVTVN2l+AH2Z5peWxwKHLKeu0c6l+UVk5PUPNNdJ5wGXAJcCP2nHUVVfBf6N5qtIC9q2AYx1E9+OwDeB22l+sH60qs5vp72P5peIW5O8bYxlx90/Q/gw8MI2bN4xUmeaSyXfpOlZWdG2UM139F9M80vjr2g+20/QdHEDzAYuT3I7zQ1jh7Sf5aNp7qW4DbgC+C5Nl/dop7bjL2jXvwR40wq0e7TjeOhn+22a4Psa8HOaruklPLQbemXacCjNPQLXA18B/r49NsfzBZpfkL89cBltWdsmzbct9vrDVY3rJODokSAeUVX30PQOHAHcTPN//z8Hpi/vmBj3eFJ38tBLKtLarT2z2qmqXrPcmSdRmru4LwM2WIUzzTVCn9qi1cNjYs3gGbx6o+3Sfz1wymTXMpY0313eIMlmwAeAc9bWH359aotWD4+JNY8Br15I85CQa4GvVtUFk13PON5A023+S5rr6n8+ueWskj61RauHx8Qaxi56SZJ6yDN4SZJ6yICXJKmHevOgmy233LJmzpw52WVIkjRhLrroot9W1ZhPBexNwM+cOZN58+ZNdhmSJE2Y9gmHY7KLXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHuo04JPMTnJlkgVJjhtnnoOTzE9yeZLT23GPSfKTJBe349/YZZ2SJPVNZ38uNskU4GRgP2AhMDfJnKqaPzDPjsDxwLOq6pYkj2on3QA8o6ruTrIxcFm77PVd1StJUp90eQa/J7Cgqq6qqnuAM4ADR81zFHByVd0CUFU3tf/eU1V3t/Ns0HGdkiT1TpfBuS1w7cDwwnbcoJ2AnZJ8P8mFSWaPTEiyXZJL2nV8wLN3SZKGN9lnxlOBHYG9gUOBjyfZFKCqrq2qXYHHAYcn2Wr0wkmOTjIvybxFixZNYNmSJK3Zugz464DtBoZntOMGLQTmVNW9VfUr4Oc0gf+A9sz9MmCv0RuoqlOqalZVzZo+ffpqLV6SpLVZlwE/F9gxyQ5J1gcOAeaMmudsmrN3kmxJ02V/VZIZSTZsx28G/AlwZYe1SpLUK50FfFUtBY4BzgOuAL5YVZcnOTHJAe1s5wGLk8wHzgeOrarFwM7Aj5L8FPgu8E9VdWlXtUqS1DepqsmuYbWYNWtWzZs3b7LL0Bpmj2NPm5DtXHTSYROyHUkalOSiqpo11rTJvslOkiR1oLMH3WjNMRFnsZ7BStKaxYCXtMquOXGXzrex/QnehiOtCANeWg0MOElrGq/BS5LUQ57BS5JW2kT0XsGa24O1JvfeeQYvSVIPeQYv9dhEPQfgK5tMyGZWmN8gWXet68c+GPCStErW5C7qiQi5NTng1nV20UuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQ50GfJLZSa5MsiDJcePMc3CS+UkuT3J6O263JD9sx12S5JVd1ilJUt9M7WrFSaYAJwP7AQuBuUnmVNX8gXl2BI4HnlVVtyR5VDvpTuCwqvpFkm2Ai5KcV1W3dlWvVs01J+4yIdvZ/oRLJ2Q7krS26/IMfk9gQVVdVVX3AGcAB46a5yjg5Kq6BaCqbmr//XlV/aJ9fz1wEzC9w1olSeqVLgN+W+DageGF7bhBOwE7Jfl+kguTzB69kiR7AusDvxxj2tFJ5iWZt2jRotVYuiRJa7fJvsluKrAjsDdwKPDxJJuOTEyyNfBZ4HVVdf/ohavqlKqaVVWzpk/3BF+SpBFdBvx1wHYDwzPacYMWAnOq6t6q+hXwc5rAJ8kjgP8B3llVF3ZYpyRJvdNlwM8FdkyyQ5L1gUOAOaPmOZvm7J0kW9J02V/Vzv8V4LSqOqvDGiVJ6qXOAr6qlgLHAOcBVwBfrKrLk5yY5IB2tvOAxUnmA+cDx1bVYuBg4NnAEUkubl+7dVWrJEl909nX5ACq6lzg3FHjThh4X8DftK/BeT4HfK7L2iRJ6rPJvslOkiR1wICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySph6ZOdgETYY9jT+t8GxeddFjn25AkaViewUuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk91GnAJ5md5MokC5IcN848ByeZn+TyJKcPjP9akluT/HeXNUqS1EedPYs+yRTgZGA/YCEwN8mcqpo/MM+OwPHAs6rqliSPGljFScBGwBu6qlGSpL7q8gx+T2BBVV1VVfcAZwAHjprnKODkqroFoKpuGplQVd8Cft9hfZIk9VaXAb8tcO3A8MJ23KCdgJ2SfD/JhUlmd1iPJEnrjMn+c7FTgR2BvYEZwAVJdqmqW4dZOMnRwNEA22+/fVc1SpK01unyDP46YLuB4RntuEELgTlVdW9V/Qr4OU3gD6WqTqmqWVU1a/r06atcsCRJfdFlwM8FdkyyQ5L1gUOAOaPmOZvm7J0kW9J02V/VYU2SJK0TOgv4qloKHAOcB1wBfLGqLk9yYpID2tnOAxYnmQ+cDxxbVYsBknwP+BKwb5KFSfbvqlZJkvqm02vwVXUucO6ocScMvC/gb9rX6GX36rI2SZL6zCfZSZLUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPdRpwCeZneTKJAuSHDfOPAcnmZ/k8iSnD4w/PMkv2tfhXdYpSVLfTO1qxUmmACcD+wELgblJ5lTV/IF5dgSOB55VVbckeVQ7fnPg74FZQAEXtcve0lW9kiT1SZdn8HsCC6rqqqq6BzgDOHDUPEcBJ48Ed1Xd1I7fH/hGVd3cTvsGMLvDWiVJ6pUuA35b4NqB4YXtuEE7ATsl+X6SC5PMXoFlSXJ0knlJ5i1atGg1li5J0tptsm+ymwrsCOwNHAp8PMmmwy5cVadU1ayqmjV9+vSOSpQkae3TZcBfB2w3MDyjHTdoITCnqu6tql8BP6cJ/GGWlSRJ4+gy4OcCOybZIcn6wCHAnFHznE1z9k6SLWm67K8CzgOen2SzJJsBz2/HSZKkIXR2F31VLU1yDE0wTwFOrarLk5wIzKuqOTwY5POB+4Bjq2oxQJJ/pPklAeDEqrq5q1olSeqbzgIeoKrOBc4dNe6EgfcF/E37Gr3sqcCpXdYnSVJfTfZNdpIkqQMGvCRJPbTcgE/ykiT+IiBJ0lpkmOB+JfCLJB9M8oSuC5IkSatuuQFfVa8BngL8Evh0kh+2T5DbpPPqJEnSShmq672qbgPOonme/NbAy4CfJHlTh7VJkqSVNMw1+AOSfAX4DrAesGdVvQB4MvDWbsuTJEkrY5jvwR8EfKiqLhgcWVV3Jnl9N2VJkqRVMUzA/wNww8hAkg2Brarq6qr6VleFSZKklTfMNfgvAfcPDN/XjpMkSWuoYQJ+alXdMzLQvl+/u5IkSdKqGibgFyU5YGQgyYHAb7srSZIkraphrsG/Efh8ko8AAa4FDuu0KkmStEqWG/BV9Uvg6Uk2bodv77wqSZK0Sob6c7FJXgT8MTAtCQBVdWKHdUmSpFUwzINuPkbzPPo30XTR/ynwmI7rkiRJq2CYm+yeWVWHAbdU1buAZwA7dVuWJElaFcME/JL23zuTbAPcS/M8ekmStIYa5hr8OUk2BU4CfgIU8PFOq5IkSatkmQGf5GHAt6rqVuDLSf4bmFZVv5uQ6iRJ0kpZZhd9Vd0PnDwwfLfhLknSmm+Ya/DfSnJQRr4fJ0mS1njDBPwbaP64zN1Jbkvy+yS3dVyXJElaBcM8yW6TiShEkiStPssN+CTPHmt8VV2w+suRJEmrwzBfkzt24P00YE/gImCfTiqSJEmrbJgu+pcMDifZDvhwZxVJkqRVNsxNdqMtBHZe3YVIkqTVZ5hr8P8fzdProPmFYDeaJ9pJkqQ11DBn8PNorrlfBPwQeEdVvWaYlSeZneTKJAuSHDfG9COSLEpycfs6cmDaB5Jc1r5eOWR7JEkSw91kdxawpKruA0gyJclGVXXnshZKMoXmKXj70XTrz00yp6rmj5r1zKo6ZtSyLwJ2p+kt2AD4TpKvVpXfv5ckaQhDPckO2HBgeEPgm0MstyewoKquqqp7gDOAA4es64nABVW1tKruAC4BZg+5rCRJ67xhAn5aVd0+MtC+32iI5bYFrh0YXtiOG+2gJJckOau9Qx/gp8DsJBsl2RJ4LrDd6AWTHJ1kXpJ5ixYtGqIkSZLWDcME/B1Jdh8ZSLIHcNdq2v45wMyq2hX4BvAZgKr6OnAu8APgCzTX/u8bvXBVnVJVs6pq1vTp01dTSZIkrf2GuQb/18CXklwPBHg0MMxNb9fx0LPuGe24B1TV4oHBTwAfHJj2HuA9AElOB34+xDYlSRLDPehmbpInAI9vR11ZVfcOse65wI5JdqAJ9kOAVw3OkGTrqrqhHTwAuKIdPwXYtKoWJ9kV2BX4+jANkiRJw30P/i+Bz1fVZe3wZkkOraqPLmu5qlqa5BjgPGAKcGpVXZ7kRGBeVc0B3pzkAGApcDNwRLv4esD32r9QexvwmqpaulItlCRpHTRMF/1RVXXyyEBV3ZLkKGCZAd/Oey7NtfTBcScMvD8eOH6M5ZbQ3EkvSZJWwjA32U1JeyoND3Sfr99dSZIkaVUNcwb/NeDMJP/RDr8B+Gp3JUmSpFU1TMC/AzgaeGM7fAnNnfSSJGkNtdwu+qq6H/gRcDXN0+n2ob3bXZIkrZnGPYNPshNwaPv6LXAmQFU9d2JKkyRJK2tZXfQ/A74HvLiqFgAkecuEVCVJklbJsrroXw7cAJyf5ONJ9qV5kp0kSVrDjRvwVXV2VR0CPAE4n+aRtY9K8u9Jnj9RBUqSpBU3zE12d1TV6VX1Eprnyf8fzZ31kiRpDTXMg24eUFW3tH/Bbd+uCpIkSatuhQJekiStHQx4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphzoN+CSzk1yZZEGS48aYfkSSRUkubl9HDkz7YJLLk1yR5N+SpMtaJUnqk6ldrTjJFOBkYD9gITA3yZyqmj9q1jOr6phRyz4TeBawazvqf4HnAN/pql5JkvqkyzP4PYEFVXVVVd0DnAEcOOSyBUwD1gc2ANYDftNJlZIk9VCXAb8tcO3A8MJ23GgHJbkkyVlJtgOoqh8C5wM3tK/zquqKDmuVJKlXJvsmu3OAmVW1K/AN4DMASR4H7AzMoPmlYJ8ke41eOMnRSeYlmbdo0aIJLFuSpDVblwF/HbDdwPCMdtwDqmpxVd3dDn4C2KN9/zLgwqq6vapuB74KPGP0BqrqlKqaVVWzpk+fvtobIEnS2qrLgJ8L7JhkhyTrA4cAcwZnSLL1wOABwEg3/DXAc5JMTbIezQ12dtFLkjSkzu6ir6qlSY4BzgOmAKdW1eVJTgTmVdUc4M1JDgCWAjcDR7SLnwXsA1xKc8Pd16rqnK5qlSSpbzoLeICqOhc4d9S4EwbeHw8cP8Zy9wFv6LI2SZL6bLJvspMkSR0w4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpB4y4CVJ6iEDXpKkHjLgJUnqIQNekqQeMuAlSeqhTgM+yewkVyZZkOS4MaYfkWRRkovb15Ht+OcOjLs4yZIkL+2yVkmS+mRqVytOMgU4GdgPWAjMTTKnquaPmvXMqjpmcERVnQ/s1q5nc2AB8PWuapUkqW+6PIPfE1hQVVdV1T3AGcCBK7GeVwBfrao7V2t1kiT1WJcBvy1w7cDwwnbcaAcluSTJWUm2G2P6IcAXxtpAkqOTzEsyb9GiRatesSRJPTHZN9mdA8ysql2BbwCfGZyYZGtgF+C8sRauqlOqalZVzZo+fXrnxUqStLboMuCvAwbPyGe04x5QVYur6u528BPAHqPWcTDwlaq6t7MqJUnqoS4Dfi6wY5IdkqxP09U+Z3CG9gx9xAHAFaPWcSjjdM9LkqTxdXYXfVUtTXIMTff6FODUqro8yYnAvKqaA7w5yQHAUuBm4IiR5ZPMpOkB+G5XNUqS1FedBTxAVZ0LnDtq3AkD748Hjh9n2asZ+6Y8SZK0HJN9k50kSeqAAS9JUg8Z8JIk9ZABL0lSDxnwkiT1kAEvSVIPGfCSJPWQAS9JUg8Z8JIk9ZABL0lSDxnwkiT1kAEvSVIPGfCSJPVQp39Nbl1yzYm7TMh2tj/h0gnZjiRp7eYZvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk9ZMBLktRDBrwkST1kwEuS1EMGvCRJPWTAS5LUQwa8JEk91GnAJ5md5MokC5IcN8b0I5IsSnJx+zpyYNr2Sb6e5Iok85PM7LJWSZL6ZGpXK04yBTgZ2A9YCMxNMqeq5o+a9cyqOmaMVZwGvKeqvpFkY+D+rmqVJKlvujyD3xNYUFVXVdU9wBnAgcMsmOSJwNSq+gZAVd1eVXd2V6okSf3SZcBvC1w7MLywHTfaQUkuSXJWku3acTsBtyb5zyT/l+SktkfgIZIcnWReknmLFi1a/S2QJGkt1VkX/ZDOAb5QVXcneQPwGWAfmrr2Ap4CXAOcCRwBfHJw4ao6BTgFYNasWTVxZUuSJtK9997LwoULWbJkyVDzn/SynTuuqPG7fLjzbVxxxRVMmzaNGTNmsN566w29XJcBfx2w3cDwjHbcA6pq8cDgJ4APtu8XAhdX1VUASc4Gns6ogJckrRsWLlzIJptswsyZM0my3Pnr2t9OQFXw2Cm/6Xwb62/9BBYvXszChQvZYYcdhl6uyy76ucCOSXZIsj5wCDBncIYkWw8MHgBcMbDspkmmt8P7AKNvzpMkrSOWLFnCFltsMVS4900Stthii6F7L0Z0dgZfVUuTHAOcB0wBTq2qy5OcCMyrqjnAm5McACwFbqbphqeq7kvyNuBbaT7Ni4CPd1WrJGnNty6G+4iVaXun34OvqnOraqeqemxVvacdd0Ib7lTV8VX1x1X15Kp6blX9bGDZb1TVrlW1S1Ud0d6JL0nSpNhl5la8fPbeHPi8vfiL172a2373OwCuv/EmDj3qLWMus98rjuCin142kWU+YLJvspMkaYXtcexpq9O6s2wAAAwRSURBVHV9n33zC5c7zwbTpvGfX/sOAMe/5S/5wmmf5Cl//Wq2efSj+MLHP7Ra61kdDHhJklbQbrs/lSt/1twadvW11/Hyw/+Sn3z7bO66awlH/83fccn8K3n843ZgyZK7H1jmU1/4Mv988qls+shN2OWJj2eD9dfnw+95J4sW38ybjjuRa6+7AYCT3vUOnvnU3Ve5RgNekqQVcN9993Hh9y/g5Ye8+g+mnXLamWy44TR++t1zuHT+lTx99sFA043/vg//Bxd+7UtssvFG7H/w69n1iY8H4K0nvJ83HXUYz9pzd6657gZe8qqj+el3z1nlOg14SZKGcPeSJbx89t7cdOON/NHjduSZe+0NPPTreP/7o4v4iz9rgn+XJz6eXXbeCYB5F1/KXk+fxeabPRKAg178fH5x1a8BOP97F/Kzn//ygXX8/vY7uP2OO9n44RutUr0GvCRJQxi5Bn/XXXdy9GsO5guf+SR/f9RQT2Bfpvvvv58LzjmdadM2WA1VPsg/FytJ0grYcMON+Nt3vZdPf/zfWbp06UOm/cnT9uDMs/8HgMt/9gsuveLnAOzx5CfxvQvnccutv2Pp0qV85dxvPrDMvs95Jh/91OcfGP7pZT9jdTDgJUlaQTs/aVd2esITOfPscx8y/ujDXskdd9zJk5/zEk78p4+w+65PBGDbrbfi7W86ij950aE896Wv5TEztuERm2wMwL/84/H85KeXM+t5L2O3vQ/g4589c7XUaBe9JGmtc9FJhy1z+vwOHlU772e/fsjwRz/1+QceVfuTb58NwIYbTuOz//5PYy5/yMtexJGv+VOWLl3Kwa//Kw6YvQ8AW26+GZ/72D+v9noNeEmSJsC7//lkvv29C1ly9z087znP4IDZ+3a6PQNekqQJ8P4Tjp3Q7XkNXpKkHjLgJUnqIQNekqQeMuAlSeohA16SpCHMesJjJruEFeJd9JKktc41J+6yzOkbr+D6bn/d+StfzBrKM3hJklbS/3z9O+z14kN52vNfwQteeSS/WdQ8YOeCH85lz/0OYs/9DuJpz38Fv7/9Dm74zSL2ffnh7LnfQey+z0v53x9dBMCZZ5/LHvu+jN33eSnvfM+/rLbaPIOXJGklPXPPp3DBOaeThFNPP4t/+ein+MDfH8uHP/ZpPvzed/LMp+7O7XfcybQN1ueTn/sS+z3nmRz3V2/gvvvu4867lnD9jTfxzvf8Cz/82hfZ7JGP4EWHHs2cr31rtTwEx4CXJGklXXfDb3jNn7+NG2/6Lffccy8zt98WgGc89Sm8410nccjLXsSBL3geM7Z5NHvs9iTe8Nb/x71Ll3LA/vvy5Cc9ge98/0c8+xlPZfoWmwNwyMtfxP9eeNFqCXi76CVJWklv+X/v5c9f9you+tZX+MgHTmDJ3XcDcOwxR/LvJ72Lu5bczXNf+lquXHAVez19Ft/88mfY5tFbcdRb3snnvvRfndZmwEuStJJuu+12tnn0owD43JfmPDD+l1dfw5N23om3/eXrmfXkJ3Hlgl/x64XXs9X0LXj9q1/B6151EBdfegWzdtuF7104j9/efAv33XcfXzz7q+z1jFmrpTa76CVJGsKSu+5inz13fWD4sKP+nL9761/wqje8lc0e+Qj2ftaeXH3tQgA+8onP8d0f/JiHPSzsvNPj2P+5e/HF//oqH/rYp1hv6lQe/vCN+OS/vpett5rOu//2Lez/p39GVfGCfZ/NS/bfZ7XUa8BLktY6259w6TKnd/HnYi/79U1/MO6xU34zZiB/6N1/+wfjXnvwgbz24AP/YPwrX/pCXvnSF66eIgfYRS9JUg8Z8JIk9ZABL0lSDxnwkqS1QlVNdgmTZmXabsBLktZ406ZNY/HixetkyFcVixcvZtq0aSu0nHfRS5LWeDNmzGDhwoUsWrRoqPlvvOX2jitq3JfbOt/G1N89jGnTpjFjxowVW66jegBIMhv4V2AK8Imqev+o6UcAJwHXtaM+UlWfaKfdB4x8D+Kaqjqgy1olSWuu9dZbjx122GHo+V9z7GkdVvOgr2xyUufbWN5XAsfTWcAnmQKcDOwHLATmJplTVfNHzXpmVR0zxiruqqrduqpPkqQ+6/Ia/J7Agqq6qqruAc4A/vAb/pIkabXrMuC3Ba4dGF7YjhvtoCSXJDkryXYD46clmZfkwiQv7bBOSZJ6J13dkZjkFcDsqjqyHX4t8LTB7vgkWwC3V9XdSd4AvLKq9mmnbVtV1yX5I+DbwL5V9ctR2zgaOLodfDxwZSeNGc6WwOp/NuLaw/bb/nW1/ety28H2T3b7H1NV08ea0OVNdtcBg2fkM3jwZjoAqmrxwOAngA8OTLuu/feqJN8BngL8ctTypwCnrNaqV1KSeVW1ev4E0FrI9tv+dbX963Lbwfavye3vsot+LrBjkh2SrA8cAswZnCHJ1gODBwBXtOM3S7JB+35L4FnA6JvzJEnSODo7g6+qpUmOAc6j+ZrcqVV1eZITgXlVNQd4c5IDgKXAzcAR7eI7A/+R5H6aX0LeP8bd95IkaRydfg++qs4Fzh017oSB98cDx4+x3A+AXbqsrQNrxKWCSWT7123rcvvX5baD7V9j29/ZTXaSJGny+Cx6SZJ6yIAfkGR2kiuTLEhy3BjTN0hyZjv9R0lmDkw7vh1/ZZL9l7fOJMe046q9kXCN0tG+ODXJTUkum5hWrLqV3Q9JtkhyfpLbk3xkouvuwhD74tlJfpJkafs12V5Z3vGbxr+1++eSJLtPdI2r01jtTbJ5km8k+UX772bjLHt4O88vkhw+cVWvmhVp87Cfd5I9klzazvdvSTJR7aGqfDWXKabQfA3vj4D1gZ8CTxw1z18AH2vfH0LzmF2AJ7bzbwDs0K5nyrLWSfO1v5nA1cCWk93+rvdFO+3ZwO7AZZPdxgnYDw8H/gR4I83fWJj09kzAvpgJ7AqcBrxismvuYB8s8/gFXgh8FQjwdOBHk13z6m4vzVeZj2vfHwd8YIzlNgeuav/drH2/2WS3Z3W3edjPG/hxOz3t/C+YqPZ4Bv+gYR6teyDwmfb9WcC+7W9jBwJnVNXdVfUrYEG7vnHXWVX/V1VXd92oldTFvqCqLqD5tsTaYqX3Q1XdUVX/CyyZuHI7tdx9UVVXV9UlwP2TUWDXhjh+DwROq8aFwKajvgq8VhmnvYPH+2eAsZ4yuj/wjaq6uapuAb4BzO6s0NVoBdu83M+7HX5EVV1YTdqfxtj7rBMG/IOGebTuA/NU1VLgd8AWy1h22Mf1rmm62Bdro1XZD33Tp8+1K+vCPtqqqm5o398IbDXGPH3bD+O1edifDwuXM09nDHhJ0gprz0jXqa9hrW1tNuAftNxH6w7Ok2Qq8Ehg8TKWHWada6Iu9sXaaFX2Q9/06XPtyrqwj34z0g3d/nvTGPP0bT+M1+Zhfz7MWM48nTHgH7TcR+u2wyN3hL4C+Hb7G90c4JD2juodgB1pbqwYZp1roi72xdpoVfZD36ytx/JEmgMc1t5d/XTgdwNdu30xeLwfDvzXGPOcBzw/zSPHNwOe345bW43X5uV+3u3wbUme3t6jdBhj77NuTPZdi2vSi+auyJ/T3C38znbcicAB7ftpwJdobhz7MfBHA8u+s13uSgbukhxrne34N9Ncj1kKXA98YrLbPwH74gvADcC9bdtfP9nt7Hg/XE1zw87tbXufONH1T/C+eGrbzjtoejEun+yaV3P7/+D4pfmWxBvb6QFObvfPpcCsya65g/ZuAXwL+AXwTWDzdt5Zgz/DgD9r/08sAF432W3pqM3jft7AxQPvZwGXtfN9hPYBcxPx8kl2kiT1kF30kiT1kAEvSVIPGfCSJPWQAS9JUg8Z8JIk9ZABL2mlJXl0kjOS/DLJRUnOTbLTZNclCaZOdgGS1k7tgzu+Anymqg5pxz2Z5lndP5/M2iQZ8JJW3nOBe6vqYyMjquqnk1iPpAF20UtaWU8CLprsIiSNzYCXJKmHDHhJK+tyYI/JLkLS2Ax4SSvr28AGSY4eGZFk1yR7TWJNkloGvKSVUs1fqnoZ8Lz2a3KXA+8DbpzcyiQB/jU5SZL6yDN4SZJ6yICXJKmHDHhJknrIgJckqYcMeEmSesiAlySphwx4SZJ6yICXJKmH/n9tr2ZEe5w6tQAAAABJRU5ErkJggg==\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["The accuracies are overall pretty similiar, besides with a very small value of C. When C is smaller, Lasso tends to do worse than Ridge. The highest accuracy was 0.629 for Ridge logistic regression, at an C = 0.01."],"metadata":{"id":"bt-fFf77epGt"}},{"cell_type":"markdown","source":["For logistic regression in SKLearn, Elastic Net regularization works with a different, and much slower solver called \"saga\". We cannot test as many values of our hyperparameters l1_ratio and C as we would like. However, since we know how the model performs on the extreme ends of the l1_ratio, we can keep these values more centralized to 0.5, as well as trying 3 values for C."],"metadata":{"id":"_JwXddDgfYQz"}},{"cell_type":"code","source":["C = [0.001, 1.0, 100.0] #inverse of regularization strength\n","l1 = [0.25, 0.5, 0.75]\n","test_acc = []\n","\n","for i in range(len(C)):\n","  for j in range(len(l1)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    logreg = LogisticRegression(max_iter=500, penalty='elasticnet', C = C[i], l1_ratio = l1[j], solver = \"saga\")\n","    logreg.fit(x_train_class, y_train_class.values.ravel())\n","\n","    # TO-DO: Find the  accuracy score and store the value\n","    score = logreg.score(x_test_class, y_test_class.values.ravel())\n","    test_acc.append(score)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"VG_-eCkvTs75","executionInfo":{"status":"ok","timestamp":1671018190498,"user_tz":300,"elapsed":1498200,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"6d2b2a02-77d9-45dc-939e-8a2f22ef1d79"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stderr","text":["/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n","/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_sag.py:352: ConvergenceWarning: The max_iter was reached which means the coef_ did not converge\n","  warnings.warn(\n"]}]},{"cell_type":"code","source":["acc = np.array(test_acc)\n","acc = acc.reshape(3,3)\n","elastic_net_df = pd.DataFrame(acc, index = C, columns = l1)\n","\n","elastic_net_df"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":143},"id":"PNrQawihTweJ","executionInfo":{"status":"ok","timestamp":1671018258737,"user_tz":300,"elapsed":111,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"158bf8c0-082b-4699-ce34-067296e94d38"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["             0.25      0.50      0.75\n","0.001    0.579197  0.579243  0.579288\n","1.000    0.579243  0.579288  0.579243\n","100.000  0.579243  0.579288  0.579243"],"text/html":["\n","  <div id=\"df-b6620a60-3b98-40c0-b9d2-66df0ccc71ae\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>0.25</th>\n","      <th>0.50</th>\n","      <th>0.75</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0.001</th>\n","      <td>0.579197</td>\n","      <td>0.579243</td>\n","      <td>0.579288</td>\n","    </tr>\n","    <tr>\n","      <th>1.000</th>\n","      <td>0.579243</td>\n","      <td>0.579288</td>\n","      <td>0.579243</td>\n","    </tr>\n","    <tr>\n","      <th>100.000</th>\n","      <td>0.579243</td>\n","      <td>0.579288</td>\n","      <td>0.579243</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b6620a60-3b98-40c0-b9d2-66df0ccc71ae')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-b6620a60-3b98-40c0-b9d2-66df0ccc71ae button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-b6620a60-3b98-40c0-b9d2-66df0ccc71ae');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":46}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(elastic_net_df, cmap = 'RdBu', vmin=0.579, vmax=0.5795)\n","plt.title(\"Heatmap for Accuracy Given Elastic Net Parameters\")\n","plt.xlabel(\"L1 Ratio\")\n","plt.ylabel(\"C\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"-UURzej1Tx9n","executionInfo":{"status":"ok","timestamp":1671019237281,"user_tz":300,"elapsed":697,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"400d2b97-4f52-4a3b-8564-edba20b12ae1"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["All the Elatic Net accuracies are extremely similar, however none of them are as good as the accuracies we were seeing with just Ridge and Lasso on their own.\n","\n","Overall, the best model based on accracy was Ridge logistic regression with C = 0.01. The accuracy of 0.629 means that with this model we were able to predict 62.9% of the test books correctly. While this accuracy is not very close to 100%, our model does have some lift. We are doing much better than classifying books at random to be above and below 4.5 mean review score."],"metadata":{"id":"DIdy4UsNgjyp"}},{"cell_type":"markdown","source":["#### 3.2.2 Random Forest"],"metadata":{"id":"VeDCp8sK4y3I"}},{"cell_type":"markdown","source":["For our second type of classification model we will try Random Forest, which is an ensemble method of Decision Trees using the idea of bagging. \n","\n","There are many parameters involved in Random Forest, however there is a consensus on the 2 most important hyperparameters to tune: n_estimators (how many decision trees in the forrest) and max_depth (maximum depth of trees)."],"metadata":{"id":"paAdsNOGoYbk"}},{"cell_type":"code","source":["from sklearn.ensemble import RandomForestClassifier\n","\n","n_estimators = [50, 100, 150, 200, 500] \n","max_depth = [5, 10, 15, 20, 25, 30]\n","test_acc = []\n","\n","for i in range(len(n_estimators)):\n","  for j in range(len(max_depth)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    rand_f = RandomForestClassifier(n_estimators = n_estimators[i], max_depth = max_depth[j])\n","    rand_f.fit(x_train_class, y_train_class.values.ravel())\n","\n","    # TO-DO: Find the  accuracy score and store the value\n","    score = rand_f.score(x_test_class, y_test_class.values.ravel())\n","    test_acc.append(score)"],"metadata":{"id":"5sqEL6e65Sop"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["acc = np.array(test_acc)\n","acc = acc.reshape(5,6)\n","rand_f_df = pd.DataFrame(acc, index = n_estimators, columns = max_depth)\n","\n","rand_f_df"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"kV4UKVt0qe-w","executionInfo":{"status":"ok","timestamp":1671022658467,"user_tz":300,"elapsed":19,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"9615d6ee-984a-49a4-f17f-729fc2355313"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["           5         10        15        20        25        30\n","50   0.609717  0.642581  0.651958  0.654394  0.656233  0.654440\n","100  0.620610  0.636468  0.648557  0.654762  0.657796  0.657106\n","150  0.617853  0.636284  0.648327  0.655727  0.659174  0.658393\n","200  0.617577  0.638628  0.652510  0.654670  0.659450  0.658485\n","500  0.618036  0.640283  0.650211  0.654992  0.657382  0.658991"],"text/html":["\n","  <div id=\"df-6360cb8c-2d87-4be5-9a81-b47657659ef3\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>5</th>\n","      <th>10</th>\n","      <th>15</th>\n","      <th>20</th>\n","      <th>25</th>\n","      <th>30</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>50</th>\n","      <td>0.609717</td>\n","      <td>0.642581</td>\n","      <td>0.651958</td>\n","      <td>0.654394</td>\n","      <td>0.656233</td>\n","      <td>0.654440</td>\n","    </tr>\n","    <tr>\n","      <th>100</th>\n","      <td>0.620610</td>\n","      <td>0.636468</td>\n","      <td>0.648557</td>\n","      <td>0.654762</td>\n","      <td>0.657796</td>\n","      <td>0.657106</td>\n","    </tr>\n","    <tr>\n","      <th>150</th>\n","      <td>0.617853</td>\n","      <td>0.636284</td>\n","      <td>0.648327</td>\n","      <td>0.655727</td>\n","      <td>0.659174</td>\n","      <td>0.658393</td>\n","    </tr>\n","    <tr>\n","      <th>200</th>\n","      <td>0.617577</td>\n","      <td>0.638628</td>\n","      <td>0.652510</td>\n","      <td>0.654670</td>\n","      <td>0.659450</td>\n","      <td>0.658485</td>\n","    </tr>\n","    <tr>\n","      <th>500</th>\n","      <td>0.618036</td>\n","      <td>0.640283</td>\n","      <td>0.650211</td>\n","      <td>0.654992</td>\n","      <td>0.657382</td>\n","      <td>0.658991</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6360cb8c-2d87-4be5-9a81-b47657659ef3')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-6360cb8c-2d87-4be5-9a81-b47657659ef3 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-6360cb8c-2d87-4be5-9a81-b47657659ef3');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":78}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(rand_f_df, cmap = 'RdBu', vmin=0.6, vmax=0.66)\n","plt.title(\"Heatmap for Accuracy Given Random Forest Parameters\")\n","plt.xlabel(\"Maximum Tree Depth\")\n","plt.ylabel(\"Number of Trees\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"eR2I36MNqfQg","executionInfo":{"status":"ok","timestamp":1671022686407,"user_tz":300,"elapsed":720,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"d969cdf3-25fb-4bb4-b4c4-4ab563c7f4a7"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Our first go around of random forest was very successful! We are seeing much higher values of accuracy than even with logisitic regression. The accuracies tend to improve with the number of trees and with the maximum tree depth. However, the best model had a max depth of 25 and 200 for the number of trees.\n","\n","We are going to tune close to these values before settling on our best random forest model."],"metadata":{"id":"jzJclEKDt4Fw"}},{"cell_type":"code","source":["n_estimators = [175, 200, 225, 250, 275, 300] \n","max_depth = [20, 25, 30, 35, 40, 45]\n","test_acc = []\n","\n","for i in range(len(n_estimators)):\n","  for j in range(len(max_depth)):\n","    # TO-DO: Initialize model with default parameters and fit it on the training set\n","    rand_f = RandomForestClassifier(n_estimators = n_estimators[i], max_depth = max_depth[j])\n","    rand_f.fit(x_train_class, y_train_class.values.ravel())\n","\n","    # TO-DO: Find the  accuracy score and store the value\n","    score = rand_f.score(x_test_class, y_test_class.values.ravel())\n","    test_acc.append(score)"],"metadata":{"id":"E_6Qj6FMtuIk"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["acc = np.array(test_acc)\n","acc = acc.reshape(6,6)\n","rand_f_df = pd.DataFrame(acc, index = n_estimators, columns = max_depth)\n","\n","rand_f_df"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":237},"id":"7BNQsFseusBa","executionInfo":{"status":"ok","timestamp":1671024754668,"user_tz":300,"elapsed":20,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"f805689a-06c4-4fb4-a878-f2ab0e523c71"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["           20        25        30        35        40        45\n","175  0.655038  0.657520  0.659037  0.659680  0.659588  0.660094\n","200  0.655589  0.656922  0.660553  0.659864  0.659220  0.658853\n","225  0.656692  0.658347  0.658393  0.659956  0.660737  0.658255\n","250  0.657198  0.658577  0.658347  0.660094  0.660507  0.659312\n","275  0.655359  0.658393  0.658669  0.660416  0.660645  0.660507\n","300  0.656141  0.657979  0.659174  0.658025  0.659910  0.660048"],"text/html":["\n","  <div id=\"df-8eaf2c3c-775f-49d8-b541-65d2c15ae88a\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>20</th>\n","      <th>25</th>\n","      <th>30</th>\n","      <th>35</th>\n","      <th>40</th>\n","      <th>45</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>175</th>\n","      <td>0.655038</td>\n","      <td>0.657520</td>\n","      <td>0.659037</td>\n","      <td>0.659680</td>\n","      <td>0.659588</td>\n","      <td>0.660094</td>\n","    </tr>\n","    <tr>\n","      <th>200</th>\n","      <td>0.655589</td>\n","      <td>0.656922</td>\n","      <td>0.660553</td>\n","      <td>0.659864</td>\n","      <td>0.659220</td>\n","      <td>0.658853</td>\n","    </tr>\n","    <tr>\n","      <th>225</th>\n","      <td>0.656692</td>\n","      <td>0.658347</td>\n","      <td>0.658393</td>\n","      <td>0.659956</td>\n","      <td>0.660737</td>\n","      <td>0.658255</td>\n","    </tr>\n","    <tr>\n","      <th>250</th>\n","      <td>0.657198</td>\n","      <td>0.658577</td>\n","      <td>0.658347</td>\n","      <td>0.660094</td>\n","      <td>0.660507</td>\n","      <td>0.659312</td>\n","    </tr>\n","    <tr>\n","      <th>275</th>\n","      <td>0.655359</td>\n","      <td>0.658393</td>\n","      <td>0.658669</td>\n","      <td>0.660416</td>\n","      <td>0.660645</td>\n","      <td>0.660507</td>\n","    </tr>\n","    <tr>\n","      <th>300</th>\n","      <td>0.656141</td>\n","      <td>0.657979</td>\n","      <td>0.659174</td>\n","      <td>0.658025</td>\n","      <td>0.659910</td>\n","      <td>0.660048</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8eaf2c3c-775f-49d8-b541-65d2c15ae88a')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-8eaf2c3c-775f-49d8-b541-65d2c15ae88a button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-8eaf2c3c-775f-49d8-b541-65d2c15ae88a');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":82}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(rand_f_df, cmap = 'RdBu', vmin=0.653, vmax=0.664)\n","plt.title(\"Heatmap for Accuracy Given Random Forest Parameters\")\n","plt.xlabel(\"Maximum Tree Depth\")\n","plt.ylabel(\"Number of Trees\")\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"wKE6T2OQusUC","executionInfo":{"status":"ok","timestamp":1671024851440,"user_tz":300,"elapsed":529,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"b7bc241b-5dab-455c-dad5-12eb8da362fb"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["With an accuracy of 0.6607 our best model is a random forest classifier with n_estimators = 225 and max_depth = 40. This model is a very strong improvement from our logistic regression model."],"metadata":{"id":"EY5OdGMluvvY"}},{"cell_type":"markdown","source":["### 3.3 Final Models on Test Set"],"metadata":{"id":"H3kpYisdA3Zo"}},{"cell_type":"markdown","source":["Are you ready for the *ultimate* test?\n","\n","Remember, we saved 10% of the data at the very start of the modelling process which was never touched until now.\n","\n","We will fit the best linear regression, logistic regression, and random forrest models on the entire training data (including the validation data we used previously). We use validation sets to avoid overfitting on training sets, but sometimes we can even overfit on those validation sets. This test set of data is completely pure, and we want to see how our final 3 models run on this data."],"metadata":{"id":"2Su9qMshc3_l"}},{"cell_type":"markdown","source":["#### 3.3.1 Linear Regression  - The Final Test"],"metadata":{"id":"C0T1l1kBT6l4"}},{"cell_type":"markdown","source":["First, we need to create the quadratic testing and training sets. "],"metadata":{"id":"_1-DIxEXivxW"}},{"cell_type":"code","source":["x_train_pow = x_train[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(2)\n","x_train_pow.columns = ['num_authors_2', 'num_categories_2', 'mean_review_percent_helpful_2', 'mean_sentiment_score_2', 'review_count_2']\n","x_train_pow = pd.concat([x_train_pow, x_train], axis=1)\n","x_test_pow = x_test_final[top_corr][['num_authors', 'num_categories', 'mean_review_percent_helpful', 'mean_sentiment_score', 'review_count']].pow(2)\n","x_test_pow.columns = ['num_authors_2', 'num_categories_2', 'mean_review_percent_helpful_2', 'mean_sentiment_score_2', 'review_count_2']\n","x_test_pow = pd.concat([x_test_pow, x_test_final], axis=1)"],"metadata":{"id":"O_5f8_49Bdzs"},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":["Then, we will train our best linear regression model (Lasso, alpha = 0.00007) and test it."],"metadata":{"id":"2qXDj1XIjFQr"}},{"cell_type":"code","source":["lasso_sq = Lasso(alpha = 0.00007)\n","lasso_sq.fit(x_train_pow, y_train['mean_review_score'])\n","\n","# TO-DO: Find the R-squared score and store the value\n","score_sq = lasso_sq.score(x_test_pow, y_test_final['mean_review_score'])\n","score_sq_train = lasso_sq.score(x_train_pow, y_train['mean_review_score'])\n","print(\"Training R-squared:\", round(score_sq_train,4))\n","print(\"Testing R-squared:\", round(score_sq,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"sTeGHPNujE-g","executionInfo":{"status":"ok","timestamp":1671020449580,"user_tz":300,"elapsed":15013,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"054892a4-f8f7-47b8-90bd-3f144cbee19d"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training R-squared: 0.1407\n","Testing R-squared: 0.1246\n"]}]},{"cell_type":"markdown","source":["Wow! Our testing $R^2$ is even higher than it was when we were validating our model. This lets us know that we did not overfit on our validation data earlier. We will call this a success!"],"metadata":{"id":"U2Y-7dz8jjSB"}},{"cell_type":"markdown","source":["#### 3.3.2 Logistic Regression - The Final Test"],"metadata":{"id":"vwIh7As-T-7Y"}},{"cell_type":"markdown","source":["Let's fit our Ridge Logistic Regression model with C = 0.01 on our entire training dataset and test it."],"metadata":{"id":"Yrnr6F6fkLUR"}},{"cell_type":"code","source":["logreg = LogisticRegression(max_iter=1000, penalty='l2', C = 0.01)\n","logreg.fit(x_train, y_train['mean_review_score_cat'].values.ravel())\n","\n","# TO-DO: Find the accuracy score and store the value\n","score = logreg.score(x_test_final, y_test_final['mean_review_score_cat'].values.ravel())\n","score_train = logreg.score(x_train, y_train['mean_review_score_cat'].values.ravel())\n","print(\"Training Accuracy:\", round(score_train,4))\n","print(\"Testing Accuracy:\", round(score,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"u0FiIMc1UEkN","executionInfo":{"status":"ok","timestamp":1671020616523,"user_tz":300,"elapsed":24881,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"5b06df2c-4209-432d-bd77-89aff4809cf4"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training Accuracy: 0.6292\n","Testing Accuracy: 0.6207\n"]}]},{"cell_type":"markdown","source":["We are pleased to see that the testing accuracy is relatively close to the training accuracy, an indication that we did not overfit. This testing accuracy is not too far off of what we were seeing while validating the model, which is good. We still are able to predict whether books have a mean review score above 4.5 or below 4.5 with more accuracy than a random guess."],"metadata":{"id":"sac-BSYYlz_x"}},{"cell_type":"markdown","source":["Let's create a confusion matrix for our testing predictions. Remember a label of 1 indicates a book with a mean review score greater than 4.5, while 0 represents a book with a mean review score less than or equal to 4.5."],"metadata":{"id":"tWSo4LsNnb_b"}},{"cell_type":"code","source":["from sklearn.metrics import confusion_matrix\n","\n","y_pred = logreg.predict(x_test_final)\n","confusion_matrix_df = pd.DataFrame(confusion_matrix(y_test_final['mean_review_score_cat'], y_pred))\n","confusion_matrix_df "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":112},"id":"Qldrqy_RmM3j","executionInfo":{"status":"ok","timestamp":1671021016042,"user_tz":300,"elapsed":123,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"4a7562f7-2990-4962-e5bd-dab19957b276"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["      0     1\n","0  5438  1432\n","1  3153  2064"],"text/html":["\n","  <div id=\"df-7a8ba1dc-5f90-4226-93bc-6f6c4227d677\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>0</th>\n","      <th>1</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>5438</td>\n","      <td>1432</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>3153</td>\n","      <td>2064</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7a8ba1dc-5f90-4226-93bc-6f6c4227d677')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-7a8ba1dc-5f90-4226-93bc-6f6c4227d677 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-7a8ba1dc-5f90-4226-93bc-6f6c4227d677');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":72}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(confusion_matrix_df, annot=True, cmap='Blues', fmt='g')\n","\n","plt.title(\"Confusion Matrix for Logistic Regression Model of Amazon Book Revies\")\n","plt.xlabel(\"Predicted Labels\")\n","plt.ylabel(\"Actual Labels\")\n","plt.show()  "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"GFqwnPy1mn4L","executionInfo":{"status":"ok","timestamp":1671021047083,"user_tz":300,"elapsed":539,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"cc671d1c-26c4-4013-d1a8-229fe7b1c20b"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["By looking at the confusion matrix, we can tell that our model predicted that a book had a label of 0 (mean review score <= 4.5) way more often than it predicted a book with a label of 1. Of the books that were actually rated 4.5+ on average, the model predicted less than half of these correctly."],"metadata":{"id":"syXplb5Hnv5A"}},{"cell_type":"markdown","source":["#### 3.3.3 Random Forest - The Final Test"],"metadata":{"id":"ab7jRbJrUGZ9"}},{"cell_type":"markdown","source":["Let's fit our best Random Forest model on the entire training dataset and test it."],"metadata":{"id":"HN6jDqnP2GGl"}},{"cell_type":"code","source":["# TO-DO: Initialize model with default parameters and fit it on the training set\n","rand_f = RandomForestClassifier(n_estimators = 225, max_depth = 40)\n","rand_f.fit(x_train, y_train['mean_review_score_cat'].values.ravel())\n","\n","# TO-DO: Find the  accuracy score and store the value\n","score = rand_f.score(x_test_final, y_test_final['mean_review_score_cat'].values.ravel())\n","score_train = rand_f.score(x_train, y_train['mean_review_score_cat'].values.ravel())\n","print(\"Training Accuracy:\", round(score_train,4))\n","print(\"Testing Accuracy:\", round(score,4))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"UhBJZZxMUF_W","executionInfo":{"status":"ok","timestamp":1671024986118,"user_tz":300,"elapsed":100368,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"a7014821-6634-4274-ef30-48f1b2c322d7"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Training Accuracy: 0.8494\n","Testing Accuracy: 0.6591\n"]}]},{"cell_type":"markdown","source":["Wow, the training accuracy is incredibly high. Many decision tree based models can overfit on the training data, but that doesn't mean they still don't do exceedingly well on the test data. This Random Forest classifier is better than the Logistic Regression classifier by almost 4% accuracy. "],"metadata":{"id":"b50t8Qp42WH6"}},{"cell_type":"code","source":["y_pred = rand_f.predict(x_test_final)\n","confusion_matrix_df = pd.DataFrame(confusion_matrix(y_test_final['mean_review_score_cat'], y_pred))\n","confusion_matrix_df "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":112},"id":"JJzr0K17r_ot","executionInfo":{"status":"ok","timestamp":1671024987110,"user_tz":300,"elapsed":995,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"cd8e5de8-62d4-4a49-ef3a-a737888015be"},"execution_count":null,"outputs":[{"output_type":"execute_result","data":{"text/plain":["      0     1\n","0  5084  1786\n","1  2335  2882"],"text/html":["\n","  <div id=\"df-a01b0418-12da-4a2f-9990-44dfd9be85d9\">\n","    <div class=\"colab-df-container\">\n","      <div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>0</th>\n","      <th>1</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>5084</td>\n","      <td>1786</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>2335</td>\n","      <td>2882</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>\n","      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a01b0418-12da-4a2f-9990-44dfd9be85d9')\"\n","              title=\"Convert this dataframe to an interactive table.\"\n","              style=\"display:none;\">\n","        \n","  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n","       width=\"24px\">\n","    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n","    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n","  </svg>\n","      </button>\n","      \n","  <style>\n","    .colab-df-container {\n","      display:flex;\n","      flex-wrap:wrap;\n","      gap: 12px;\n","    }\n","\n","    .colab-df-convert {\n","      background-color: #E8F0FE;\n","      border: none;\n","      border-radius: 50%;\n","      cursor: pointer;\n","      display: none;\n","      fill: #1967D2;\n","      height: 32px;\n","      padding: 0 0 0 0;\n","      width: 32px;\n","    }\n","\n","    .colab-df-convert:hover {\n","      background-color: #E2EBFA;\n","      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n","      fill: #174EA6;\n","    }\n","\n","    [theme=dark] .colab-df-convert {\n","      background-color: #3B4455;\n","      fill: #D2E3FC;\n","    }\n","\n","    [theme=dark] .colab-df-convert:hover {\n","      background-color: #434B5C;\n","      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n","      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n","      fill: #FFFFFF;\n","    }\n","  </style>\n","\n","      <script>\n","        const buttonEl =\n","          document.querySelector('#df-a01b0418-12da-4a2f-9990-44dfd9be85d9 button.colab-df-convert');\n","        buttonEl.style.display =\n","          google.colab.kernel.accessAllowed ? 'block' : 'none';\n","\n","        async function convertToInteractive(key) {\n","          const element = document.querySelector('#df-a01b0418-12da-4a2f-9990-44dfd9be85d9');\n","          const dataTable =\n","            await google.colab.kernel.invokeFunction('convertToInteractive',\n","                                                     [key], {});\n","          if (!dataTable) return;\n","\n","          const docLinkHtml = 'Like what you see? Visit the ' +\n","            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n","            + ' to learn more about interactive tables.';\n","          element.innerHTML = '';\n","          dataTable['output_type'] = 'display_data';\n","          await google.colab.output.renderOutput(dataTable, element);\n","          const docLink = document.createElement('div');\n","          docLink.innerHTML = docLinkHtml;\n","          element.appendChild(docLink);\n","        }\n","      </script>\n","    </div>\n","  </div>\n","  "]},"metadata":{},"execution_count":89}]},{"cell_type":"code","source":["f, ax = plt.subplots(figsize=(8, 6))\n","sns.heatmap(confusion_matrix_df, annot=True, cmap='Blues', fmt='g')\n","\n","plt.title(\"Confusion Matrix for Random Forest Model of Amazon Book Revies\")\n","plt.xlabel(\"Predicted Labels\")\n","plt.ylabel(\"Actual Labels\")\n","plt.show()  "],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":404},"id":"F0Rhd0yNsEpv","executionInfo":{"status":"ok","timestamp":1671024987259,"user_tz":300,"elapsed":155,"user":{"displayName":"Grace Benner","userId":"05774739731217404352"}},"outputId":"7405e69c-9030-458a-d24c-7ede41ddf889"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 576x432 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Finally, examining the confusion matrix, and especially after comparing it to the Logistic Regression's, we can be confident this is the better classification model all around. Not only does it have a higher accuracy, but the diagonals (correctly predicted books) contain more observations than the off diagonals. Our results are more balanced.\n","\n","We trust this RF classifier to predict a book correctly (whether or not the mean review score is above 4.5) 65.9% of the time."],"metadata":{"id":"WMF_xLOv3Aw1"}},{"cell_type":"markdown","source":["## Conclusion"],"metadata":{"id":"3dQEcn0K5cZB"}},{"cell_type":"markdown","metadata":{"id":"oQNYaTMspbWT"},"source":["# THE END!\n","\n","Thank you for following along on our incredible journey! It is our sincere hope that you can now find a book whose pages you can lose yourself in, and finally say that you did do your 'reading'! Happy holidays! \n","- Grace, Kofi, Naomi \n","\n","\n","<br>\n","<center><img src = \"https://ih1.redbubble.net/image.1016783569.8606/pp,504x498-pad,600x600,f8f8f8.jpg\" width= \"500\" height='400' align =\"center\"/></center>\n","<br>"]}]}