weather_guru.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "weather-guru.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyPXhdnior6K+WSwxCoAne52",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/marlonrcfranco/weather-guru/blob/main/weather_guru.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RJ0V8yZcc1g9"
      },
      "source": [
        "# weather-guru\n",
        "## Will it rain tomorrow?\n",
        "\n",
        "\n",
        "### Goal 🌦\n",
        "Implement an algorithm that performs **next day rain prediction** by training machine learning models on the target variable `RainTomorrow`.\n",
        "\n",
        "\n",
        "\n",
        "### Dataset 📂\n",
        "The dataset contains about **10 years of daily weather observations** from various locations in **Australia**.\n",
        "\n",
        "**`RainTomorrow`** is the target variable to be  predicted. It means - it rained the next day, this column is `Yes` if the rain that day was 1mm or more.\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F80l3nZwiSEj"
      },
      "source": [
        "## Imports & Defines"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hpvWyhLNimyM",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "edb5ee2a-7db3-476c-f1b3-fdd587a4b851"
      },
      "source": [
        "%matplotlib inline\n",
        "import re\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.ticker as mticker\n",
        "import matplotlib.patches as mpatches\n",
        "\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.keras.wrappers.scikit_learn import KerasClassifier\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.preprocessing import LabelEncoder,MinMaxScaler,StandardScaler\n",
        "from sklearn.model_selection import StratifiedKFold\n",
        "from sklearn.impute import SimpleImputer\n",
        "from sklearn.pipeline import Pipeline\n",
        "\n",
        "from geopy.geocoders import Nominatim\n",
        "from geopy.extra.rate_limiter import RateLimiter\n",
        "from functools import partial\n",
        "import plotly.express as px\n",
        "!pip install portolan\n",
        "import portolan\n",
        "\n",
        "# Defines\n",
        "DATASET_URL = \"https://raw.githubusercontent.com/marlonrcfranco/weather-guru/main/data/weatherAUS.csv\"\n",
        "TARGET_NAME = 'RainTomorrow'\n",
        "\n",
        "EPOCHS = 100\n",
        "# The number of epochs is the number of complete passes through the training dataset. \n",
        "# The number 100 is arbitrary, chosen to improve the model accuracy and training time.\n",
        "\n",
        "BATCH_SIZE = 1023\n",
        "# The size of a batch must be more than or equal to one and less than or equal to the number of samples in the training dataset.\n",
        "# The number 1023 is also arbitrary, chosen to improve the model accuracy and training time.\n",
        "\n",
        "K_FOLD_SPLITS = 5\n",
        "# The number of iterations used in the cross validation. \n",
        "# The number 5 is arbitrary, chosen to reduce the time the validation takes.\n",
        "\n",
        "LOSS_FUNCTION = 'binary_crossentropy'\n",
        "# Binary cross entropy compares each of the predicted probabilities to actual class output which can be either 0 or 1. \n",
        "# The loss function 'binary_crossentropy' was chosen since the target variable is binary.\n",
        "\n",
        "LEARNING_RATE = 0.001\n",
        "# Learning rate used in the Adam optimizer. \n",
        "# The value of 0.001 is arbitrary, chosen by trial and error, evaluating the model's improvements.\n",
        "\n",
        "NEURONS_1 = 27\n",
        "# The number of neurons in the first hidden layer. \n",
        "# The values 27 is arbitrary, chosen by trial and error, evaluating the model's improvements.\n",
        "\n",
        "NEURONS_2 = 7\n",
        "# The number of neurons in the first hidden layer. \n",
        "# The values 7 was chosen to 'force' the model to choose the most relevant features\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: portolan in /usr/local/lib/python3.7/dist-packages (1.0.1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XhowXUCfjM5r"
      },
      "source": [
        "## Fetch Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eqPoSDMxjYH5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "eb985ce2-99c4-4943-8f12-1906dfecd5fb"
      },
      "source": [
        "weather_df = pd.read_csv(DATASET_URL,parse_dates=True)\n",
        "weather_df.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(145460, 23)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G5Nz4VmGkvvi"
      },
      "source": [
        "## Data Preprocessing"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-irldH4cnZHq"
      },
      "source": [
        "### Dataframe Info"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U7GpWG_lk07u",
        "outputId": "dc037ed4-04a5-4496-f811-a4d269557530"
      },
      "source": [
        "weather_df.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 145460 entries, 0 to 145459\n",
            "Data columns (total 23 columns):\n",
            " #   Column         Non-Null Count   Dtype  \n",
            "---  ------         --------------   -----  \n",
            " 0   Date           145460 non-null  object \n",
            " 1   Location       145460 non-null  object \n",
            " 2   MinTemp        143975 non-null  float64\n",
            " 3   MaxTemp        144199 non-null  float64\n",
            " 4   Rainfall       142199 non-null  float64\n",
            " 5   Evaporation    82670 non-null   float64\n",
            " 6   Sunshine       75625 non-null   float64\n",
            " 7   WindGustDir    135134 non-null  object \n",
            " 8   WindGustSpeed  135197 non-null  float64\n",
            " 9   WindDir9am     134894 non-null  object \n",
            " 10  WindDir3pm     141232 non-null  object \n",
            " 11  WindSpeed9am   143693 non-null  float64\n",
            " 12  WindSpeed3pm   142398 non-null  float64\n",
            " 13  Humidity9am    142806 non-null  float64\n",
            " 14  Humidity3pm    140953 non-null  float64\n",
            " 15  Pressure9am    130395 non-null  float64\n",
            " 16  Pressure3pm    130432 non-null  float64\n",
            " 17  Cloud9am       89572 non-null   float64\n",
            " 18  Cloud3pm       86102 non-null   float64\n",
            " 19  Temp9am        143693 non-null  float64\n",
            " 20  Temp3pm        141851 non-null  float64\n",
            " 21  RainToday      142199 non-null  object \n",
            " 22  RainTomorrow   142193 non-null  object \n",
            "dtypes: float64(16), object(7)\n",
            "memory usage: 25.5+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XjEPTRAKnKbU"
      },
      "source": [
        "### Sample"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 117
        },
        "id": "a30TK-tGmTNA",
        "outputId": "8d861834-f422-410d-d329-d4b76fe9b92a"
      },
      "source": [
        "weather_df.sample()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>Date</th>\n",
              "      <th>Location</th>\n",
              "      <th>MinTemp</th>\n",
              "      <th>MaxTemp</th>\n",
              "      <th>Rainfall</th>\n",
              "      <th>Evaporation</th>\n",
              "      <th>Sunshine</th>\n",
              "      <th>WindGustDir</th>\n",
              "      <th>WindGustSpeed</th>\n",
              "      <th>WindDir9am</th>\n",
              "      <th>WindDir3pm</th>\n",
              "      <th>WindSpeed9am</th>\n",
              "      <th>WindSpeed3pm</th>\n",
              "      <th>Humidity9am</th>\n",
              "      <th>Humidity3pm</th>\n",
              "      <th>Pressure9am</th>\n",
              "      <th>Pressure3pm</th>\n",
              "      <th>Cloud9am</th>\n",
              "      <th>Cloud3pm</th>\n",
              "      <th>Temp9am</th>\n",
              "      <th>Temp3pm</th>\n",
              "      <th>RainToday</th>\n",
              "      <th>RainTomorrow</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>57625</th>\n",
              "      <td>2016-01-26</td>\n",
              "      <td>Ballarat</td>\n",
              "      <td>13.5</td>\n",
              "      <td>27.4</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>SE</td>\n",
              "      <td>57.0</td>\n",
              "      <td>SE</td>\n",
              "      <td>SE</td>\n",
              "      <td>43.0</td>\n",
              "      <td>22.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>33.0</td>\n",
              "      <td>1016.9</td>\n",
              "      <td>1014.8</td>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>26.2</td>\n",
              "      <td>No</td>\n",
              "      <td>No</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "             Date  Location  MinTemp  ...  Temp3pm  RainToday  RainTomorrow\n",
              "57625  2016-01-26  Ballarat     13.5  ...     26.2         No            No\n",
              "\n",
              "[1 rows x 23 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JwFHZtOqvkxU"
      },
      "source": [
        "### Unique values for non-numeric columns"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qJ8cJHKrO0dT"
      },
      "source": [
        "There are some 'string' values in the dataset. Let's see what these values are."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JLNKQxEOvxBG",
        "outputId": "02440c23-bd49-4231-f380-ff394be27937"
      },
      "source": [
        "for col in weather_df:\n",
        "  if not pd.api.types.is_numeric_dtype(weather_df[col]):\n",
        "    print(\"\\n-------\\nColumn: \"+col+\" [\"+str(len(weather_df[col].unique()))+\" unique values]:\\n\")\n",
        "    print(weather_df[col].unique())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "-------\n",
            "Column: Date [3436 unique values]:\n",
            "\n",
            "['2008-12-01' '2008-12-02' '2008-12-03' ... '2008-01-29' '2008-01-30'\n",
            " '2008-01-31']\n",
            "\n",
            "-------\n",
            "Column: Location [49 unique values]:\n",
            "\n",
            "['Albury' 'BadgerysCreek' 'Cobar' 'CoffsHarbour' 'Moree' 'Newcastle'\n",
            " 'NorahHead' 'NorfolkIsland' 'Penrith' 'Richmond' 'Sydney' 'SydneyAirport'\n",
            " 'WaggaWagga' 'Williamtown' 'Wollongong' 'Canberra' 'Tuggeranong'\n",
            " 'MountGinini' 'Ballarat' 'Bendigo' 'Sale' 'MelbourneAirport' 'Melbourne'\n",
            " 'Mildura' 'Nhil' 'Portland' 'Watsonia' 'Dartmoor' 'Brisbane' 'Cairns'\n",
            " 'GoldCoast' 'Townsville' 'Adelaide' 'MountGambier' 'Nuriootpa' 'Woomera'\n",
            " 'Albany' 'Witchcliffe' 'PearceRAAF' 'PerthAirport' 'Perth' 'SalmonGums'\n",
            " 'Walpole' 'Hobart' 'Launceston' 'AliceSprings' 'Darwin' 'Katherine'\n",
            " 'Uluru']\n",
            "\n",
            "-------\n",
            "Column: WindGustDir [17 unique values]:\n",
            "\n",
            "['W' 'WNW' 'WSW' 'NE' 'NNW' 'N' 'NNE' 'SW' nan 'ENE' 'SSE' 'S' 'NW' 'SE'\n",
            " 'ESE' 'E' 'SSW']\n",
            "\n",
            "-------\n",
            "Column: WindDir9am [17 unique values]:\n",
            "\n",
            "['W' 'NNW' 'SE' 'ENE' 'SW' 'SSE' 'S' 'NE' nan 'SSW' 'N' 'WSW' 'ESE' 'E'\n",
            " 'NW' 'WNW' 'NNE']\n",
            "\n",
            "-------\n",
            "Column: WindDir3pm [17 unique values]:\n",
            "\n",
            "['WNW' 'WSW' 'E' 'NW' 'W' 'SSE' 'ESE' 'ENE' 'NNW' 'SSW' 'SW' 'SE' 'N' 'S'\n",
            " 'NNE' nan 'NE']\n",
            "\n",
            "-------\n",
            "Column: RainToday [3 unique values]:\n",
            "\n",
            "['No' 'Yes' nan]\n",
            "\n",
            "-------\n",
            "Column: RainTomorrow [3 unique values]:\n",
            "\n",
            "['No' 'Yes' nan]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uNAlhF_nW00m"
      },
      "source": [
        "So there are 3436 days of observation in the dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2QxWYwWkR-1L"
      },
      "source": [
        "### Check for duplicated rows"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UWZFBm81R-O2",
        "outputId": "37902c9e-1a5e-4b0b-f453-a60608b138d9"
      },
      "source": [
        "# Check if exists duplicated rows based on all columns\n",
        "duplicateRowsDF = weather_df[weather_df.duplicated()]\n",
        "if len(duplicateRowsDF)!=0: print(\"Identical rows were found!\")\n",
        "\n",
        "# Check if exists is duplicated rows based on 'Date' column\n",
        "duplicateRowsDF = weather_df[weather_df.duplicated(['Date'])]\n",
        "if len(duplicateRowsDF)!=0: \n",
        "  print(len(duplicateRowsDF),\"rows with the same Date were found.\\n\")\n",
        "\n",
        "# Check if exists is duplicated rows based on 'Date' and 'Location' columns\n",
        "duplicateRowsDF = weather_df[weather_df.duplicated(['Date','Location'])]\n",
        "if len(duplicateRowsDF)!=0: \n",
        "  print(len(duplicateRowsDF),\"rows with the same Date and Location were found.\\n\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "142024 rows with the same Date were found.\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pE04PyDszMUE"
      },
      "source": [
        "### Convert Date to datetime"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0HUTy5eLQO5K"
      },
      "source": [
        "Since we have a column called 'Date' in a date-like format, let's convert it to datetime."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I2SMRC11zRk6"
      },
      "source": [
        "weather_df['Date'] = pd.to_datetime(weather_df['Date'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FOrX2IccWd_a"
      },
      "source": [
        "#### Wind Cardinal directions (string) to Degrees (float)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AwpNnuvFQXLW"
      },
      "source": [
        "Wind directions are represented in Cardinal directions (e.g. N, S, SW,...). In order to make the dataset only with numbers, let's convert these directions into angles (degrees)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        },
        "id": "OdkZ5gbZRBWS",
        "outputId": "9d4b58e2-fada-4892-b644-db5f8d02f0e2"
      },
      "source": [
        "weather_df[['WindGustDir','WindDir9am','WindDir3pm']].head(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>WindGustDir</th>\n",
              "      <th>WindDir9am</th>\n",
              "      <th>WindDir3pm</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>W</td>\n",
              "      <td>W</td>\n",
              "      <td>WNW</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  WindGustDir WindDir9am WindDir3pm\n",
              "0           W          W        WNW"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O6jHeV_aW2-A"
      },
      "source": [
        "weather_df['WindGustDir'] = weather_df['WindGustDir'].apply(lambda w: portolan.middle(str(w)) if str(w)!='nan' else w)\n",
        "weather_df['WindDir9am'] = weather_df['WindDir9am'].apply(lambda w: portolan.middle(str(w)) if str(w)!='nan' else w)\n",
        "weather_df['WindDir3pm'] = weather_df['WindDir3pm'].apply(lambda w: portolan.middle(str(w)) if str(w)!='nan' else w)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        },
        "id": "iFh1A45bQu-T",
        "outputId": "68c6ea18-9426-4eea-8bca-2cff9a19f9fb"
      },
      "source": [
        "weather_df[['WindGustDir','WindDir9am','WindDir3pm']].head(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>WindGustDir</th>\n",
              "      <th>WindDir9am</th>\n",
              "      <th>WindDir3pm</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>270.0</td>\n",
              "      <td>270.0</td>\n",
              "      <td>292.5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   WindGustDir  WindDir9am  WindDir3pm\n",
              "0        270.0       270.0       292.5"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Savqoh3XW3fV"
      },
      "source": [
        "### Map 'Yes' and 'No' to infer boolean values\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VJUyk8KOW9vr"
      },
      "source": [
        "weather_df.RainToday = weather_df.RainToday.map(dict(Yes=1, No=0))\n",
        "weather_df.RainTomorrow = weather_df.RainTomorrow.map(dict(Yes=1, No=0))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gGnZy4YdvU6i"
      },
      "source": [
        "### Add YEAR, MONTH and DAY columns"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZRaElrBbvpDH"
      },
      "source": [
        "weather_df['year'] = weather_df.Date.dt.year\n",
        "weather_df['month'] = weather_df.Date.dt.month\n",
        "weather_df['day'] = weather_df.Date.dt.day\n",
        "\n",
        "# Drop the Date column\n",
        "# It is now replaced by year, month and day columns\n",
        "weather_df.drop(['Date'], axis=1, inplace=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XOGrY_-_Fkoo"
      },
      "source": [
        "### Add latitude and longitude columns"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3YmbA2S2RWaI"
      },
      "source": [
        "In order to catch cyclic behaviors, we can feed the model with the day, month and year separately and allow it to perceive patterns related to seasons, for example."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R_gki7EEFjLl",
        "outputId": "e91a2bc1-b572-42e3-86f8-664332d117a6"
      },
      "source": [
        "locator = Nominatim(user_agent=\"weather-guru\")\n",
        "# conveneint function to delay between geocoding calls\n",
        "geocode = RateLimiter(locator.geocode, min_delay_seconds=1)\n",
        "\n",
        "# 1 - List the unique values of Location, to further improve performance\n",
        "unique_locations = pd.DataFrame()\n",
        "unique_locations['description'] = weather_df['Location'].unique()\n",
        "unique_locations['id'] = weather_df['Location'].unique()\n",
        "unique_locations.set_index(['id'], inplace = True)\n",
        "print(\"Total unique locations: {}\".format(len(unique_locations)))\n",
        "\n",
        "# Add 'rainfall/day' column\n",
        "unique_locations['days'] = weather_df['Location'].value_counts().values\n",
        "unique_locations['rainfall'] = weather_df.groupby(['Location']).sum()['Rainfall'].values\n",
        "unique_locations['rainfall/day'] = unique_locations['rainfall']/unique_locations['days']\n",
        "\n",
        "# Separate CamelCase with space to allow geocode to find the location\n",
        "unique_locations['description'] = unique_locations['description'].apply(lambda loc: re.sub('([A-Z][a-z]+)', r' \\1', re.sub('([A-Z]+)', r' \\1', loc)))\n",
        "# Add the country to the Location in order to avoid misunderstandings\n",
        "unique_locations['description'] = unique_locations['description'].apply(lambda loc: loc + ', Australia')\n",
        "\n",
        "# 2- create improved location column (gcode)\n",
        "print(\"Searching for locations...\")\n",
        "unique_locations['gcode'] =  unique_locations['description'].apply(partial(geocode, addressdetails=True))\n",
        "\n",
        "# 3 - create point (longitude, latitude and altitude) from gcode (location) column (returns tuple)\n",
        "unique_locations['point'] = unique_locations['gcode'].apply(lambda loc: tuple(loc.point) if loc else None)\n",
        "\n",
        "# 4 - split point column into latitude, longitude and altitude columns\n",
        "unique_locations[['latitude', 'longitude', 'altitude']] = pd.DataFrame(unique_locations['point'].tolist(), index=unique_locations.index)\n",
        "\n",
        "# Verify if all the locations were found\n",
        "if not unique_locations['gcode'].isnull().any(): \n",
        "  print(\"All the locations were found!\")\n",
        "else: \n",
        "  print(\"The current locations were not found:\")\n",
        "  print(unique_locations[unique_locations['gcode'].isnull()])\n",
        "\n",
        "# Verify if all the found locations are from Australia\n",
        "unique_locations['country'] = unique_locations['gcode'].apply(lambda loc: (loc.raw.get('address').get('country')) if loc else None)\n",
        "if unique_locations['country'].str.contains('Australia').all():\n",
        "  print(\"All the found locations are from Australia\")\n",
        "\n",
        "def get_lat_lon_alt_from_location(location_name):\n",
        "  location_data = unique_locations.loc[location_name]\n",
        "  lat = location_data['latitude']\n",
        "  lon = location_data['longitude']\n",
        "  alt = location_data['altitude']\n",
        "  return lat,lon,alt\n",
        "\n",
        "print(\"Adding latitude, longitude and altitude columns to the dataset...\")\n",
        "weather_df[['latitude', 'longitude', 'altitude']] = list(weather_df['Location'].apply(get_lat_lon_alt_from_location))\n",
        "\n",
        "# Drop the Location column\n",
        "# It is now replaced by latitude, longitude and altitude columns\n",
        "weather_df.drop(['Location'], axis=1, inplace=True)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total unique locations: 49\n",
            "Searching for locations...\n",
            "All the locations were found!\n",
            "All the found locations are from Australia\n",
            "Adding latitude, longitude and altitude columns to the dataset...\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "osFzfVl9jG6P",
        "outputId": "7da1d226-5488-494b-a8ea-d77b5e174818"
      },
      "source": [
        "weather_df.altitude.unique()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hj8Tz-kdjCVk"
      },
      "source": [
        "The `altitude` column is 0 for all the observations, so we can remove it from the dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zoK5un8IjBZD"
      },
      "source": [
        "weather_df.drop('altitude',axis=1,inplace=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g2G3uSI5dUey"
      },
      "source": [
        "### Plot map (just for fun)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 542
        },
        "id": "GTUMsvLtgNRK",
        "outputId": "1cd321d8-0a93-43a2-d1ab-3857d92dfba4"
      },
      "source": [
        "# Plot map\n",
        "fig = px.scatter_geo(unique_locations,lat='latitude',lon='longitude', hover_name=\"description\", color='rainfall/day')\n",
        "fig.update_layout(title = 'Locations in the dataset', title_x=0.5)\n",
        "fig.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>\n",
              "            <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>\n",
              "                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-latest.min.js\"></script>    \n",
              "            <div id=\"64f295b0-58de-4947-8d94-f7f6f425ebf4\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>\n",
              "            <script type=\"text/javascript\">\n",
              "                \n",
              "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
              "                    \n",
              "                if (document.getElementById(\"64f295b0-58de-4947-8d94-f7f6f425ebf4\")) {\n",
              "                    Plotly.newPlot(\n",
              "                        '64f295b0-58de-4947-8d94-f7f6f425ebf4',\n",
              "                        [{\"geo\": \"geo\", \"hoverlabel\": {\"namelength\": 0}, \"hovertemplate\": \"<b>%{hovertext}</b><br><br>latitude=%{lat}<br>longitude=%{lon}<br>rainfall/day=%{marker.color}\", \"hovertext\": [\"  Albury, Australia\", \"  Badgerys  Creek, Australia\", \"  Cobar, Australia\", \"  Coffs  Harbour, Australia\", \"  Moree, Australia\", \"  Newcastle, Australia\", \"  Norah  Head, Australia\", \"  Norfolk  Island, Australia\", \"  Penrith, Australia\", \"  Richmond, Australia\", \"  Sydney, Australia\", \"  Sydney  Airport, Australia\", \"  Wagga  Wagga, Australia\", \"  Williamtown, Australia\", \"  Wollongong, Australia\", \"  Canberra, Australia\", \"  Tuggeranong, Australia\", \"  Mount  Ginini, Australia\", \"  Ballarat, Australia\", \"  Bendigo, Australia\", \"  Sale, Australia\", \"  Melbourne  Airport, Australia\", \"  Melbourne, Australia\", \"  Mildura, Australia\", \"  Nhil, Australia\", \"  Portland, Australia\", \"  Watsonia, Australia\", \"  Dartmoor, Australia\", \"  Brisbane, Australia\", \"  Cairns, Australia\", \"  Gold  Coast, Australia\", \"  Townsville, Australia\", \"  Adelaide, Australia\", \"  Mount  Gambier, Australia\", \"  Nuriootpa, Australia\", \"  Woomera, Australia\", \"  Albany, Australia\", \"  Witchcliffe, Australia\", \"  Pearce RAAF, Australia\", \"  Perth  Airport, Australia\", \"  Perth, Australia\", \"  Salmon  Gums, Australia\", \"  Walpole, Australia\", \"  Hobart, Australia\", \"  Launceston, Australia\", \"  Alice  Springs, Australia\", \"  Darwin, Australia\", \"  Katherine, Australia\", \"  Uluru, Australia\"], \"lat\": [-36.0804766, -33.8816671, -31.4983333, -30.2962407, -29.4617202, -32.9192953, -33.2816667, -29.0289575, -33.7510792, -20.569655, -33.8548157, -33.9498935, -35.115, -32.815, -34.4243941, -35.2975906, -35.4209771, -35.5297196, -37.5622632, -36.7588767, -38.1050358, -37.667111000000006, -37.8142176, -34.1847265, -35.4325283, -38.3456231, -37.7110022, -37.9225444, -27.4689682, -16.9206657, -28.0023731, -19.2569391, -34.9281805, -37.8246698, -34.4693354, -31.1999142, -35.0247822, -34.0263348, -31.6739604, -31.9406095, -31.9527121, -32.9815347, -34.9776796, -42.8825088, -41.4340813, -23.6983884, -12.46044, -14.4642313, -25.3455545], \"legendgroup\": \"\", \"lon\": [146.9162795, 150.7441627, 145.8344444, 153.1135293, 149.8407153, 151.7795348, 151.5677778, 167.9587289126371, 150.6941684, 142.9283602143393, 151.2164539, 151.18196819346016, 147.3677778, 151.8427778, 150.89385, 149.1012676, 149.0921341, 148.7726345, 143.8607215, 144.2825931, 147.0647902, 144.83348076679553, 144.9631608, 142.1624972, 141.2833187, 141.6042304, 145.083635, 141.2766551, 153.0234991, 145.7721854, 153.4145987, 146.8239537, 138.5999312, 140.7820067624113, 138.9939006, 136.8253532, 117.883608, 115.1004768, 116.01754351808195, 115.96760765137932, 115.8604796, 121.6439417, 116.7310063, 147.3281233, 147.1373496, 133.8812885, 130.8410469, 132.2638568, 131.03696147470208], \"marker\": {\"color\": [1.409080325960418, 2.0418062200956872, 1.8050109614782262, 0.8383338553084874, 2.011086752270583, 1.650109614782324, 1.5387409959285834, 3.1133730034450404, 5.643815789473699, 1.9582894736842016, 1.1080263157894723, 4.916644736842121, 2.0773684210526193, 5.348750000000006, 3.6950000000000105, 1.6798026315789405, 1.6426644736842104, 2.0040460526315753, 1.4978947368421027, 1.4371710526315757, 0.9351102336294838, 1.5309641329384647, 2.082066469233286, 3.180658025922229, 3.1267530741110043, 0.4874709205716201, 3.2972416085078127, 3.0808906613492804, 1.3875706214689234, 1.5320704553007614, 2.142771684945152, 2.022864739115981, 1.7616483881688256, 2.519441674975064, 2.097241608507798, 1.505649717514116, 1.015819209039546, 3.686939182452635, 3.0059155865736154, 3.5133931538717187, 2.1561316051844432, 0.39674310402126967, 1.6911930874044487, 2.726014637391873, 1.8577230359520582, 3.055014995001665, 5.416983523447395, 6.793409378960716, 0.9295310519645137], \"coloraxis\": \"coloraxis\"}, \"name\": \"\", \"showlegend\": false, \"type\": \"scattergeo\"}],\n",
              "                        {\"coloraxis\": {\"colorbar\": {\"title\": {\"text\": \"rainfall/day\"}}, \"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\"]]}, \"geo\": {\"center\": {}, \"domain\": {\"x\": [0.0, 1.0], \"y\": [0.0, 1.0]}}, \"legend\": {\"tracegroupgap\": 0}, \"margin\": {\"t\": 60}, \"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"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\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"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}, \"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}}}, \"title\": {\"text\": \"Locations in the dataset\", \"x\": 0.5}},\n",
              "                        {\"responsive\": true}\n",
              "                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('64f295b0-58de-4947-8d94-f7f6f425ebf4');\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",
              "                        })\n",
              "                };\n",
              "                \n",
              "            </script>\n",
              "        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sgzTY1tyT4Nx"
      },
      "source": [
        "![aus_rain_observations.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bDHSkiBaXPM2"
      },
      "source": [
        "### Reorder columns"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LPAMLeuARhXh"
      },
      "source": [
        "Put the date and coordinates columns into the beggining, and the target column at the end."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2kq0L3IhTx_G"
      },
      "source": [
        "# Reorder columns\n",
        "initial_columns = ['year','month','day','latitude','longitude']\n",
        "target_column = [TARGET_NAME]\n",
        "middle_columns = [col for col in weather_df.columns if col not in initial_columns and col not in target_column]\n",
        "columns_reordered = initial_columns + middle_columns + target_column\n",
        "\n",
        "weather_df =  weather_df[columns_reordered]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r909dR3tW_gG"
      },
      "source": [
        "### Proceed with data type conversion"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cj8xgLJbW8Yg"
      },
      "source": [
        "# Convert data types\n",
        "weather_df = weather_df.convert_dtypes()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lXdct1vWuqUk"
      },
      "source": [
        "## Data Cleaning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LkyUBpXfn_-K"
      },
      "source": [
        "### Remove rows with null values in the target column"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qJ_STnSERrE_"
      },
      "source": [
        "If there's no value in the target column, we cannot use it into our training or test set. So we can remove those rows from the dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a0LuiPD0oXZp",
        "outputId": "dfcc746f-521a-466a-9ae9-a782ffbbfdba"
      },
      "source": [
        "# There are null values (nan) in the RainTomorrow column\n",
        "print(\"Before:\\n  Unique values in the RainTomorrow column: \", weather_df.RainTomorrow.unique())\n",
        "rows_before = len(weather_df)\n",
        "print(\" Total number of rows: [\", rows_before, \"]\")\n",
        "# Select only the not-null rows\n",
        "weather_df = weather_df[weather_df.RainTomorrow.notna()]\n",
        "print(\"\\nAfter:\\n  Unique values in the RainTomorrow column: \", weather_df.RainTomorrow.unique())\n",
        "rows_after = len(weather_df)\n",
        "print(\" Total number of rows: [\", rows_after, \"] (\", (rows_before-rows_after) , \" rows removed )\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Before:\n",
            "  Unique values in the RainTomorrow column:  <IntegerArray>\n",
            "[0, 1, <NA>]\n",
            "Length: 3, dtype: Int64\n",
            " Total number of rows: [ 145460 ]\n",
            "\n",
            "After:\n",
            "  Unique values in the RainTomorrow column:  <IntegerArray>\n",
            "[0, 1]\n",
            "Length: 2, dtype: Int64\n",
            " Total number of rows: [ 142193 ] ( 3267  rows removed )\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aLV0A4-Q-mXA"
      },
      "source": [
        "### Ensure all data is float"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Wvhw-qvJR-4j"
      },
      "source": [
        "It will help with the filling method later."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-qBP3-JdSWWZ"
      },
      "source": [
        "for col in weather_df.columns:\n",
        "  weather_df[col] = weather_df[col].astype('float32')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L57w3hL2Jbz_"
      },
      "source": [
        "### Filling missing values "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SHsWp5EbSGk7"
      },
      "source": [
        "In this step, we use the 'mean' strategy to fill the missing values of the features."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "93HCf9dkNk9w"
      },
      "source": [
        "imputer = SimpleImputer(missing_values=np.nan, strategy='mean')\n",
        "# transform the dataset\n",
        "weather_df_transformed = pd.DataFrame(imputer.fit_transform(weather_df))\n",
        "weather_df_transformed.columns = weather_df.columns\n",
        "weather_df_transformed.index = weather_df.index\n",
        "\n",
        "weather_df = weather_df_transformed"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RPyoYdgh4CzG"
      },
      "source": [
        "### Correlation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 833
        },
        "id": "IUVz16Q44E2y",
        "outputId": "ae854f90-851e-4a47-c58d-f347ee07c90d"
      },
      "source": [
        "f = plt.figure(figsize=(20, 17))\n",
        "plt.matshow(weather_df.corr(), fignum=f.number)\n",
        "plt.xticks(range(weather_df.select_dtypes(['number']).shape[1]), weather_df.select_dtypes(['number']).columns, fontsize=14, rotation=90)\n",
        "plt.yticks(range(weather_df.select_dtypes(['number']).shape[1]), weather_df.select_dtypes(['number']).columns, fontsize=14)\n",
        "cb = plt.colorbar()\n",
        "cb.ax.tick_params(labelsize=17)\n",
        "plt.title('Correlation Matrix', fontsize=16, pad=90);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x1224 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3_-v211kUdwf"
      },
      "source": [
        "There are some correlation in the dataset. Enough to proceed with the work."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dU_jnEU8ghfy"
      },
      "source": [
        "## Feature engineering"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sCV0JY-di8TM"
      },
      "source": [
        "values = weather_df.values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7e7rcAkJkp_y",
        "outputId": "7557b3b5-9a9d-4d60-b92a-e113d9830165"
      },
      "source": [
        "# Number of features is the number of columns except the target column\n",
        "NUM_FEATURES = len(weather_df.columns)-1\n",
        "NUM_FEATURES"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "25"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "keMPn3eh78VI"
      },
      "source": [
        "### Normalize features"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cqW3U1fnSgVT"
      },
      "source": [
        "The MinMaxScaler puts the values between 0 and 1, which further improves the model performance."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J4o16e7y8BNq"
      },
      "source": [
        "scaler = MinMaxScaler(feature_range=(0, 1))\n",
        "values = scaler.fit_transform(values)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EuLEeNLg_5PA"
      },
      "source": [
        "### Split into input and outputs"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o01Aenk5Srwh"
      },
      "source": [
        "X = features\n",
        "\n",
        "Y = target"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UvhtBiFvpel0",
        "outputId": "ff8cd1fa-cbd9-4c85-ca21-c904956706e8"
      },
      "source": [
        "target_col_idx = weather_df.columns.get_loc(TARGET_NAME)\n",
        "\n",
        "X = values[:,0:target_col_idx].astype(float)\n",
        "Y = values[:,target_col_idx]\n",
        "\n",
        "print(X.shape,Y.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(142193, 25) (142193,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HlwU8pZArI2u"
      },
      "source": [
        "### Encode class values as integers"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DfQKdsxbrJcW"
      },
      "source": [
        "encoder = LabelEncoder()\n",
        "encoder.fit(Y)\n",
        "encoded_Y = encoder.transform(Y)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WVV-QmR-59Cq"
      },
      "source": [
        "## Design the model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rZRpXyXt2xjz"
      },
      "source": [
        "We want to predict a boolean value ('Yes' or 'No') for the target variable **`RainTomorrow`**. In this case, we need to use a **classification** model, istead of a **regression** model, wich is used to predict real-world values (e.g. `Rainfall`)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RZ_bl0cp6DMg"
      },
      "source": [
        "def create_model():\n",
        "  model = Sequential(name='weather_guru')\n",
        "  model.add(Dense(NEURONS_1, input_dim=NUM_FEATURES, activation='relu', name='dense_1'))\n",
        "  model.add(Dense(NEURONS_2, activation='relu', name='dense_2'))\n",
        "  model.add(Dense(1, activation='sigmoid', name='output'))\n",
        "  adam_optmz = Adam(learning_rate=LEARNING_RATE)\n",
        "  # Compile model\n",
        "  model.compile(loss=LOSS_FUNCTION, optimizer=adam_optmz, metrics=['accuracy'])\n",
        "  print(model.summary())\n",
        "  return model\n",
        "\n",
        "estimators = []\n",
        "estimators.append(('standardize', StandardScaler()))\n",
        "estimators.append(('mlp', KerasClassifier(build_fn = create_model, epochs = EPOCHS, batch_size = BATCH_SIZE, verbose = 1)))\n",
        "pipeline = Pipeline(estimators)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GkBzihGe6Dvo"
      },
      "source": [
        "## Train the model and validate it"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r_LQI3L2S3Fn"
      },
      "source": [
        "We use the cross validation score to measure the accuracy of our model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yudgAVlyKgdn"
      },
      "source": [
        "kfold = StratifiedKFold(n_splits=K_FOLD_SPLITS, shuffle=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NSMNDHibTC61"
      },
      "source": [
        "The kfold will repeat the training `K_FOLD_SPLITS` times, selecting different sets of **train** and **test** data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1KDdEfxCoRRS",
        "outputId": "963a400b-4bff-4fcb-a826-d737b9883f20"
      },
      "source": [
        "%%time\n",
        "results = cross_val_score(pipeline, X, encoded_Y, cv=kfold)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"weather_guru\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 27)                702       \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 7)                 196       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 1)                 8         \n",
            "=================================================================\n",
            "Total params: 906\n",
            "Trainable params: 906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "112/112 [==============================] - 1s 2ms/step - loss: 0.5105 - accuracy: 0.7862\n",
            "Epoch 2/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3808 - accuracy: 0.8381\n",
            "Epoch 3/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3648 - accuracy: 0.8433\n",
            "Epoch 4/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3576 - accuracy: 0.8456\n",
            "Epoch 5/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3531 - accuracy: 0.8471\n",
            "Epoch 6/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3499 - accuracy: 0.8485\n",
            "Epoch 7/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3476 - accuracy: 0.8497\n",
            "Epoch 8/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3456 - accuracy: 0.8504\n",
            "Epoch 9/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3442 - accuracy: 0.8511\n",
            "Epoch 10/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3427 - accuracy: 0.8521\n",
            "Epoch 11/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3418 - accuracy: 0.8525\n",
            "Epoch 12/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3406 - accuracy: 0.8528\n",
            "Epoch 13/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3400 - accuracy: 0.8532\n",
            "Epoch 14/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3392 - accuracy: 0.8533\n",
            "Epoch 15/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3384 - accuracy: 0.8533\n",
            "Epoch 16/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3379 - accuracy: 0.8536\n",
            "Epoch 17/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3373 - accuracy: 0.8539\n",
            "Epoch 18/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3369 - accuracy: 0.8544\n",
            "Epoch 19/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3362 - accuracy: 0.8543\n",
            "Epoch 20/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3357 - accuracy: 0.8544\n",
            "Epoch 21/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3354 - accuracy: 0.8547\n",
            "Epoch 22/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3351 - accuracy: 0.8550\n",
            "Epoch 23/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3345 - accuracy: 0.8550\n",
            "Epoch 24/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3343 - accuracy: 0.8550\n",
            "Epoch 25/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3339 - accuracy: 0.8548\n",
            "Epoch 26/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3334 - accuracy: 0.8554\n",
            "Epoch 27/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3332 - accuracy: 0.8555\n",
            "Epoch 28/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3328 - accuracy: 0.8549\n",
            "Epoch 29/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3325 - accuracy: 0.8556\n",
            "Epoch 30/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3321 - accuracy: 0.8560\n",
            "Epoch 31/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3319 - accuracy: 0.8561\n",
            "Epoch 32/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3315 - accuracy: 0.8557\n",
            "Epoch 33/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3312 - accuracy: 0.8560\n",
            "Epoch 34/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3309 - accuracy: 0.8557\n",
            "Epoch 35/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3308 - accuracy: 0.8560\n",
            "Epoch 36/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3304 - accuracy: 0.8564\n",
            "Epoch 37/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3302 - accuracy: 0.8561\n",
            "Epoch 38/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3300 - accuracy: 0.8568\n",
            "Epoch 39/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3297 - accuracy: 0.8561\n",
            "Epoch 40/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3296 - accuracy: 0.8562\n",
            "Epoch 41/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3291 - accuracy: 0.8570\n",
            "Epoch 42/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3290 - accuracy: 0.8569\n",
            "Epoch 43/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3286 - accuracy: 0.8571\n",
            "Epoch 44/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3286 - accuracy: 0.8572\n",
            "Epoch 45/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3282 - accuracy: 0.8572\n",
            "Epoch 46/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3281 - accuracy: 0.8575\n",
            "Epoch 47/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3279 - accuracy: 0.8572\n",
            "Epoch 48/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3276 - accuracy: 0.8577\n",
            "Epoch 49/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3276 - accuracy: 0.8573\n",
            "Epoch 50/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3272 - accuracy: 0.8583\n",
            "Epoch 51/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3272 - accuracy: 0.8578\n",
            "Epoch 52/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3271 - accuracy: 0.8579\n",
            "Epoch 53/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3268 - accuracy: 0.8583\n",
            "Epoch 54/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3267 - accuracy: 0.8579\n",
            "Epoch 55/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3266 - accuracy: 0.8583\n",
            "Epoch 56/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3264 - accuracy: 0.8581\n",
            "Epoch 57/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3262 - accuracy: 0.8580\n",
            "Epoch 58/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3262 - accuracy: 0.8584\n",
            "Epoch 59/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3260 - accuracy: 0.8582\n",
            "Epoch 60/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3259 - accuracy: 0.8582\n",
            "Epoch 61/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3258 - accuracy: 0.8586\n",
            "Epoch 62/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3257 - accuracy: 0.8587\n",
            "Epoch 63/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3254 - accuracy: 0.8587\n",
            "Epoch 64/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3254 - accuracy: 0.8585\n",
            "Epoch 65/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3251 - accuracy: 0.8586\n",
            "Epoch 66/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3251 - accuracy: 0.8587\n",
            "Epoch 67/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3250 - accuracy: 0.8587\n",
            "Epoch 68/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8591\n",
            "Epoch 69/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8587\n",
            "Epoch 70/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3247 - accuracy: 0.8589\n",
            "Epoch 71/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3245 - accuracy: 0.8596\n",
            "Epoch 72/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3244 - accuracy: 0.8591\n",
            "Epoch 73/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8596\n",
            "Epoch 74/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8595\n",
            "Epoch 75/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8593\n",
            "Epoch 76/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8591\n",
            "Epoch 77/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3239 - accuracy: 0.8598\n",
            "Epoch 78/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3238 - accuracy: 0.8593\n",
            "Epoch 79/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3238 - accuracy: 0.8597\n",
            "Epoch 80/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3235 - accuracy: 0.8595\n",
            "Epoch 81/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3236 - accuracy: 0.8598\n",
            "Epoch 82/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3236 - accuracy: 0.8598\n",
            "Epoch 83/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3234 - accuracy: 0.8598\n",
            "Epoch 84/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8602\n",
            "Epoch 85/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3233 - accuracy: 0.8599\n",
            "Epoch 86/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3233 - accuracy: 0.8601\n",
            "Epoch 87/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8596\n",
            "Epoch 88/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3231 - accuracy: 0.8601\n",
            "Epoch 89/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3231 - accuracy: 0.8601\n",
            "Epoch 90/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8601\n",
            "Epoch 91/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8603\n",
            "Epoch 92/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8598\n",
            "Epoch 93/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8603\n",
            "Epoch 94/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3227 - accuracy: 0.8601\n",
            "Epoch 95/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8603\n",
            "Epoch 96/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8604\n",
            "Epoch 97/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3226 - accuracy: 0.8604\n",
            "Epoch 98/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8602\n",
            "Epoch 99/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8604\n",
            "Epoch 100/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8603\n",
            "28/28 [==============================] - 0s 1ms/step - loss: 0.3282 - accuracy: 0.8573\n",
            "Model: \"weather_guru\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 27)                702       \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 7)                 196       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 1)                 8         \n",
            "=================================================================\n",
            "Total params: 906\n",
            "Trainable params: 906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "112/112 [==============================] - 1s 2ms/step - loss: 0.5883 - accuracy: 0.6866\n",
            "Epoch 2/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3945 - accuracy: 0.8290\n",
            "Epoch 3/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3687 - accuracy: 0.8396\n",
            "Epoch 4/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3597 - accuracy: 0.8436\n",
            "Epoch 5/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3547 - accuracy: 0.8459\n",
            "Epoch 6/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3511 - accuracy: 0.8471\n",
            "Epoch 7/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3482 - accuracy: 0.8485\n",
            "Epoch 8/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3458 - accuracy: 0.8497\n",
            "Epoch 9/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3438 - accuracy: 0.8507\n",
            "Epoch 10/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3423 - accuracy: 0.8517\n",
            "Epoch 11/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3408 - accuracy: 0.8526\n",
            "Epoch 12/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3396 - accuracy: 0.8530\n",
            "Epoch 13/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3385 - accuracy: 0.8536\n",
            "Epoch 14/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3375 - accuracy: 0.8539\n",
            "Epoch 15/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3365 - accuracy: 0.8544\n",
            "Epoch 16/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3359 - accuracy: 0.8547\n",
            "Epoch 17/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3352 - accuracy: 0.8552\n",
            "Epoch 18/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3346 - accuracy: 0.8553\n",
            "Epoch 19/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3340 - accuracy: 0.8555\n",
            "Epoch 20/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3335 - accuracy: 0.8557\n",
            "Epoch 21/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3331 - accuracy: 0.8561\n",
            "Epoch 22/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3326 - accuracy: 0.8562\n",
            "Epoch 23/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3323 - accuracy: 0.8561\n",
            "Epoch 24/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3319 - accuracy: 0.8565\n",
            "Epoch 25/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3316 - accuracy: 0.8564\n",
            "Epoch 26/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3311 - accuracy: 0.8568\n",
            "Epoch 27/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3307 - accuracy: 0.8568\n",
            "Epoch 28/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3306 - accuracy: 0.8567\n",
            "Epoch 29/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3302 - accuracy: 0.8570\n",
            "Epoch 30/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3297 - accuracy: 0.8574\n",
            "Epoch 31/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3296 - accuracy: 0.8572\n",
            "Epoch 32/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3292 - accuracy: 0.8574\n",
            "Epoch 33/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3289 - accuracy: 0.8578\n",
            "Epoch 34/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3289 - accuracy: 0.8576\n",
            "Epoch 35/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3287 - accuracy: 0.8573\n",
            "Epoch 36/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3284 - accuracy: 0.8580\n",
            "Epoch 37/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3281 - accuracy: 0.8579\n",
            "Epoch 38/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3278 - accuracy: 0.8579\n",
            "Epoch 39/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3277 - accuracy: 0.8583\n",
            "Epoch 40/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3276 - accuracy: 0.8581\n",
            "Epoch 41/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3274 - accuracy: 0.8584\n",
            "Epoch 42/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3271 - accuracy: 0.8585\n",
            "Epoch 43/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3270 - accuracy: 0.8592\n",
            "Epoch 44/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3269 - accuracy: 0.8586\n",
            "Epoch 45/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3266 - accuracy: 0.8587\n",
            "Epoch 46/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3265 - accuracy: 0.8589\n",
            "Epoch 47/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3264 - accuracy: 0.8591\n",
            "Epoch 48/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3263 - accuracy: 0.8591\n",
            "Epoch 49/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3260 - accuracy: 0.8592\n",
            "Epoch 50/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3259 - accuracy: 0.8595\n",
            "Epoch 51/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3258 - accuracy: 0.8596\n",
            "Epoch 52/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3257 - accuracy: 0.8593\n",
            "Epoch 53/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3256 - accuracy: 0.8593\n",
            "Epoch 54/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3255 - accuracy: 0.8599\n",
            "Epoch 55/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3253 - accuracy: 0.8594\n",
            "Epoch 56/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3252 - accuracy: 0.8597\n",
            "Epoch 57/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3250 - accuracy: 0.8597\n",
            "Epoch 58/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8599\n",
            "Epoch 59/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8595\n",
            "Epoch 60/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3246 - accuracy: 0.8600\n",
            "Epoch 61/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8596\n",
            "Epoch 62/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3244 - accuracy: 0.8597\n",
            "Epoch 63/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3244 - accuracy: 0.8605\n",
            "Epoch 64/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8597\n",
            "Epoch 65/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8603\n",
            "Epoch 66/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3239 - accuracy: 0.8602\n",
            "Epoch 67/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3238 - accuracy: 0.8599\n",
            "Epoch 68/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3237 - accuracy: 0.8604\n",
            "Epoch 69/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3236 - accuracy: 0.8603\n",
            "Epoch 70/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3234 - accuracy: 0.8605\n",
            "Epoch 71/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3235 - accuracy: 0.8602\n",
            "Epoch 72/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3235 - accuracy: 0.8602\n",
            "Epoch 73/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3233 - accuracy: 0.8607\n",
            "Epoch 74/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8604\n",
            "Epoch 75/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3230 - accuracy: 0.8604\n",
            "Epoch 76/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3230 - accuracy: 0.8607\n",
            "Epoch 77/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8607\n",
            "Epoch 78/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8606\n",
            "Epoch 79/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3226 - accuracy: 0.8605\n",
            "Epoch 80/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8608\n",
            "Epoch 81/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8607\n",
            "Epoch 82/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8611\n",
            "Epoch 83/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3223 - accuracy: 0.8610\n",
            "Epoch 84/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3222 - accuracy: 0.8607\n",
            "Epoch 85/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3222 - accuracy: 0.8612\n",
            "Epoch 86/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3220 - accuracy: 0.8607\n",
            "Epoch 87/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8611\n",
            "Epoch 88/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3220 - accuracy: 0.8613\n",
            "Epoch 89/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3218 - accuracy: 0.8613\n",
            "Epoch 90/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3215 - accuracy: 0.8614\n",
            "Epoch 91/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3216 - accuracy: 0.8610\n",
            "Epoch 92/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3215 - accuracy: 0.8614\n",
            "Epoch 93/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3214 - accuracy: 0.8614\n",
            "Epoch 94/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8614\n",
            "Epoch 95/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8614\n",
            "Epoch 96/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8613\n",
            "Epoch 97/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8614\n",
            "Epoch 98/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8617\n",
            "Epoch 99/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3211 - accuracy: 0.8616\n",
            "Epoch 100/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3209 - accuracy: 0.8615\n",
            "28/28 [==============================] - 0s 1ms/step - loss: 0.3317 - accuracy: 0.8552\n",
            "Model: \"weather_guru\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 27)                702       \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 7)                 196       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 1)                 8         \n",
            "=================================================================\n",
            "Total params: 906\n",
            "Trainable params: 906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "112/112 [==============================] - 1s 2ms/step - loss: 0.4905 - accuracy: 0.7678\n",
            "Epoch 2/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3816 - accuracy: 0.8331\n",
            "Epoch 3/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3664 - accuracy: 0.8402\n",
            "Epoch 4/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3591 - accuracy: 0.8439\n",
            "Epoch 5/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3544 - accuracy: 0.8465\n",
            "Epoch 6/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3509 - accuracy: 0.8487\n",
            "Epoch 7/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3480 - accuracy: 0.8498\n",
            "Epoch 8/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3456 - accuracy: 0.8509\n",
            "Epoch 9/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3436 - accuracy: 0.8521\n",
            "Epoch 10/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3420 - accuracy: 0.8525\n",
            "Epoch 11/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3405 - accuracy: 0.8534\n",
            "Epoch 12/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3394 - accuracy: 0.8535\n",
            "Epoch 13/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3382 - accuracy: 0.8543\n",
            "Epoch 14/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3372 - accuracy: 0.8546\n",
            "Epoch 15/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3364 - accuracy: 0.8549\n",
            "Epoch 16/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3357 - accuracy: 0.8554\n",
            "Epoch 17/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3350 - accuracy: 0.8556\n",
            "Epoch 18/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3344 - accuracy: 0.8558\n",
            "Epoch 19/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3339 - accuracy: 0.8560\n",
            "Epoch 20/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3334 - accuracy: 0.8563\n",
            "Epoch 21/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3329 - accuracy: 0.8563\n",
            "Epoch 22/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3325 - accuracy: 0.8570\n",
            "Epoch 23/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3321 - accuracy: 0.8568\n",
            "Epoch 24/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3317 - accuracy: 0.8570\n",
            "Epoch 25/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3313 - accuracy: 0.8576\n",
            "Epoch 26/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3311 - accuracy: 0.8576\n",
            "Epoch 27/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3307 - accuracy: 0.8574\n",
            "Epoch 28/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3303 - accuracy: 0.8579\n",
            "Epoch 29/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3299 - accuracy: 0.8581\n",
            "Epoch 30/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3296 - accuracy: 0.8578\n",
            "Epoch 31/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3293 - accuracy: 0.8582\n",
            "Epoch 32/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3291 - accuracy: 0.8580\n",
            "Epoch 33/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3288 - accuracy: 0.8584\n",
            "Epoch 34/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3286 - accuracy: 0.8583\n",
            "Epoch 35/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3283 - accuracy: 0.8584\n",
            "Epoch 36/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3281 - accuracy: 0.8588\n",
            "Epoch 37/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3278 - accuracy: 0.8591\n",
            "Epoch 38/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3277 - accuracy: 0.8589\n",
            "Epoch 39/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3274 - accuracy: 0.8593\n",
            "Epoch 40/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3274 - accuracy: 0.8590\n",
            "Epoch 41/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3271 - accuracy: 0.8592\n",
            "Epoch 42/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3269 - accuracy: 0.8594\n",
            "Epoch 43/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3268 - accuracy: 0.8588\n",
            "Epoch 44/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3266 - accuracy: 0.8593\n",
            "Epoch 45/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3263 - accuracy: 0.8593\n",
            "Epoch 46/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3262 - accuracy: 0.8598\n",
            "Epoch 47/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3260 - accuracy: 0.8594\n",
            "Epoch 48/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3259 - accuracy: 0.8595\n",
            "Epoch 49/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3259 - accuracy: 0.8595\n",
            "Epoch 50/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3256 - accuracy: 0.8598\n",
            "Epoch 51/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3257 - accuracy: 0.8597\n",
            "Epoch 52/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3254 - accuracy: 0.8601\n",
            "Epoch 53/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3252 - accuracy: 0.8600\n",
            "Epoch 54/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3251 - accuracy: 0.8599\n",
            "Epoch 55/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3249 - accuracy: 0.8600\n",
            "Epoch 56/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8602\n",
            "Epoch 57/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3248 - accuracy: 0.8601\n",
            "Epoch 58/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3245 - accuracy: 0.8605\n",
            "Epoch 59/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3245 - accuracy: 0.8601\n",
            "Epoch 60/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8599\n",
            "Epoch 61/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3247 - accuracy: 0.8602\n",
            "Epoch 62/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3241 - accuracy: 0.8606\n",
            "Epoch 63/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3241 - accuracy: 0.8608\n",
            "Epoch 64/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3237 - accuracy: 0.8607\n",
            "Epoch 65/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3237 - accuracy: 0.8603\n",
            "Epoch 66/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3239 - accuracy: 0.8610\n",
            "Epoch 67/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3234 - accuracy: 0.8610\n",
            "Epoch 68/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3234 - accuracy: 0.8608\n",
            "Epoch 69/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3233 - accuracy: 0.8612\n",
            "Epoch 70/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8607\n",
            "Epoch 71/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3230 - accuracy: 0.8610\n",
            "Epoch 72/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8615\n",
            "Epoch 73/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8609\n",
            "Epoch 74/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8612\n",
            "Epoch 75/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3227 - accuracy: 0.8613\n",
            "Epoch 76/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3226 - accuracy: 0.8615\n",
            "Epoch 77/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8610\n",
            "Epoch 78/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3223 - accuracy: 0.8613\n",
            "Epoch 79/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3223 - accuracy: 0.8616\n",
            "Epoch 80/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3221 - accuracy: 0.8616\n",
            "Epoch 81/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3219 - accuracy: 0.8615\n",
            "Epoch 82/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8615\n",
            "Epoch 83/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3217 - accuracy: 0.8615\n",
            "Epoch 84/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3217 - accuracy: 0.8619\n",
            "Epoch 85/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3215 - accuracy: 0.8619\n",
            "Epoch 86/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3215 - accuracy: 0.8617\n",
            "Epoch 87/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8613\n",
            "Epoch 88/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3214 - accuracy: 0.8617\n",
            "Epoch 89/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3211 - accuracy: 0.8618\n",
            "Epoch 90/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8613\n",
            "Epoch 91/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8618\n",
            "Epoch 92/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8619\n",
            "Epoch 93/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8619\n",
            "Epoch 94/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3208 - accuracy: 0.8621\n",
            "Epoch 95/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3208 - accuracy: 0.8619\n",
            "Epoch 96/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3208 - accuracy: 0.8619\n",
            "Epoch 97/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3205 - accuracy: 0.8621\n",
            "Epoch 98/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3204 - accuracy: 0.8623\n",
            "Epoch 99/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3204 - accuracy: 0.8626\n",
            "Epoch 100/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3205 - accuracy: 0.8620\n",
            "28/28 [==============================] - 0s 1ms/step - loss: 0.3281 - accuracy: 0.8588\n",
            "Model: \"weather_guru\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 27)                702       \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 7)                 196       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 1)                 8         \n",
            "=================================================================\n",
            "Total params: 906\n",
            "Trainable params: 906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "112/112 [==============================] - 1s 2ms/step - loss: 0.4831 - accuracy: 0.7680\n",
            "Epoch 2/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3859 - accuracy: 0.8308\n",
            "Epoch 3/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3692 - accuracy: 0.8385\n",
            "Epoch 4/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3603 - accuracy: 0.8436\n",
            "Epoch 5/100\n",
            "112/112 [==============================] - 0s 1ms/step - loss: 0.3552 - accuracy: 0.8462\n",
            "Epoch 6/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3515 - accuracy: 0.8477\n",
            "Epoch 7/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3488 - accuracy: 0.8494\n",
            "Epoch 8/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3466 - accuracy: 0.8503\n",
            "Epoch 9/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3446 - accuracy: 0.8513\n",
            "Epoch 10/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3429 - accuracy: 0.8521\n",
            "Epoch 11/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3413 - accuracy: 0.8527\n",
            "Epoch 12/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3396 - accuracy: 0.8535\n",
            "Epoch 13/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3384 - accuracy: 0.8544\n",
            "Epoch 14/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3373 - accuracy: 0.8546\n",
            "Epoch 15/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3363 - accuracy: 0.8553\n",
            "Epoch 16/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3354 - accuracy: 0.8553\n",
            "Epoch 17/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3346 - accuracy: 0.8559\n",
            "Epoch 18/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3340 - accuracy: 0.8560\n",
            "Epoch 19/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3334 - accuracy: 0.8569\n",
            "Epoch 20/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3329 - accuracy: 0.8567\n",
            "Epoch 21/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3324 - accuracy: 0.8561\n",
            "Epoch 22/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3318 - accuracy: 0.8569\n",
            "Epoch 23/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3314 - accuracy: 0.8571\n",
            "Epoch 24/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3310 - accuracy: 0.8569\n",
            "Epoch 25/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3305 - accuracy: 0.8571\n",
            "Epoch 26/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3304 - accuracy: 0.8574\n",
            "Epoch 27/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3296 - accuracy: 0.8576\n",
            "Epoch 28/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3293 - accuracy: 0.8578\n",
            "Epoch 29/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3289 - accuracy: 0.8573\n",
            "Epoch 30/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3286 - accuracy: 0.8579\n",
            "Epoch 31/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3283 - accuracy: 0.8581\n",
            "Epoch 32/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3280 - accuracy: 0.8579\n",
            "Epoch 33/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3278 - accuracy: 0.8579\n",
            "Epoch 34/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3274 - accuracy: 0.8586\n",
            "Epoch 35/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3271 - accuracy: 0.8583\n",
            "Epoch 36/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3267 - accuracy: 0.8582\n",
            "Epoch 37/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3267 - accuracy: 0.8581\n",
            "Epoch 38/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3262 - accuracy: 0.8581\n",
            "Epoch 39/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3259 - accuracy: 0.8588\n",
            "Epoch 40/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3256 - accuracy: 0.8590\n",
            "Epoch 41/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3254 - accuracy: 0.8590\n",
            "Epoch 42/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3251 - accuracy: 0.8587\n",
            "Epoch 43/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3251 - accuracy: 0.8588\n",
            "Epoch 44/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3249 - accuracy: 0.8588\n",
            "Epoch 45/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3246 - accuracy: 0.8593\n",
            "Epoch 46/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3244 - accuracy: 0.8590\n",
            "Epoch 47/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3242 - accuracy: 0.8593\n",
            "Epoch 48/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3241 - accuracy: 0.8588\n",
            "Epoch 49/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3238 - accuracy: 0.8594\n",
            "Epoch 50/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3236 - accuracy: 0.8594\n",
            "Epoch 51/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3235 - accuracy: 0.8595\n",
            "Epoch 52/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3234 - accuracy: 0.8596\n",
            "Epoch 53/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8596\n",
            "Epoch 54/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3231 - accuracy: 0.8597\n",
            "Epoch 55/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8599\n",
            "Epoch 56/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3227 - accuracy: 0.8594\n",
            "Epoch 57/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8599\n",
            "Epoch 58/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8595\n",
            "Epoch 59/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3222 - accuracy: 0.8602\n",
            "Epoch 60/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8603\n",
            "Epoch 61/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3220 - accuracy: 0.8603\n",
            "Epoch 62/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8602\n",
            "Epoch 63/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3218 - accuracy: 0.8604\n",
            "Epoch 64/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3216 - accuracy: 0.8603\n",
            "Epoch 65/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3217 - accuracy: 0.8602\n",
            "Epoch 66/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3214 - accuracy: 0.8605\n",
            "Epoch 67/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8608\n",
            "Epoch 68/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8605\n",
            "Epoch 69/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8605\n",
            "Epoch 70/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8607\n",
            "Epoch 71/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8606\n",
            "Epoch 72/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3207 - accuracy: 0.8606\n",
            "Epoch 73/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3207 - accuracy: 0.8607\n",
            "Epoch 74/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3206 - accuracy: 0.8605\n",
            "Epoch 75/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3206 - accuracy: 0.8613\n",
            "Epoch 76/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3204 - accuracy: 0.8608\n",
            "Epoch 77/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3204 - accuracy: 0.8611\n",
            "Epoch 78/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3203 - accuracy: 0.8610\n",
            "Epoch 79/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3202 - accuracy: 0.8611\n",
            "Epoch 80/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3201 - accuracy: 0.8609\n",
            "Epoch 81/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3200 - accuracy: 0.8611\n",
            "Epoch 82/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3200 - accuracy: 0.8610\n",
            "Epoch 83/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3200 - accuracy: 0.8612\n",
            "Epoch 84/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3197 - accuracy: 0.8611\n",
            "Epoch 85/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3198 - accuracy: 0.8610\n",
            "Epoch 86/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3197 - accuracy: 0.8615\n",
            "Epoch 87/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3197 - accuracy: 0.8616\n",
            "Epoch 88/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3193 - accuracy: 0.8613\n",
            "Epoch 89/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3195 - accuracy: 0.8614\n",
            "Epoch 90/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3195 - accuracy: 0.8609\n",
            "Epoch 91/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3192 - accuracy: 0.8615\n",
            "Epoch 92/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3192 - accuracy: 0.8617\n",
            "Epoch 93/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3192 - accuracy: 0.8614\n",
            "Epoch 94/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3190 - accuracy: 0.8614\n",
            "Epoch 95/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3190 - accuracy: 0.8612\n",
            "Epoch 96/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3189 - accuracy: 0.8618\n",
            "Epoch 97/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3190 - accuracy: 0.8610\n",
            "Epoch 98/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3188 - accuracy: 0.8616\n",
            "Epoch 99/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3186 - accuracy: 0.8622\n",
            "Epoch 100/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3187 - accuracy: 0.8618\n",
            "28/28 [==============================] - 0s 1ms/step - loss: 0.3302 - accuracy: 0.8579\n",
            "Model: \"weather_guru\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "dense_1 (Dense)              (None, 27)                702       \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 7)                 196       \n",
            "_________________________________________________________________\n",
            "output (Dense)               (None, 1)                 8         \n",
            "=================================================================\n",
            "Total params: 906\n",
            "Trainable params: 906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n",
            "Epoch 1/100\n",
            "112/112 [==============================] - 1s 2ms/step - loss: 0.4545 - accuracy: 0.7940\n",
            "Epoch 2/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3774 - accuracy: 0.8356\n",
            "Epoch 3/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3624 - accuracy: 0.8430\n",
            "Epoch 4/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3551 - accuracy: 0.8468\n",
            "Epoch 5/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3504 - accuracy: 0.8490\n",
            "Epoch 6/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3473 - accuracy: 0.8499\n",
            "Epoch 7/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3448 - accuracy: 0.8514\n",
            "Epoch 8/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3430 - accuracy: 0.8523\n",
            "Epoch 9/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3413 - accuracy: 0.8529\n",
            "Epoch 10/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3402 - accuracy: 0.8532\n",
            "Epoch 11/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3390 - accuracy: 0.8541\n",
            "Epoch 12/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3381 - accuracy: 0.8547\n",
            "Epoch 13/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3374 - accuracy: 0.8545\n",
            "Epoch 14/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3366 - accuracy: 0.8548\n",
            "Epoch 15/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3358 - accuracy: 0.8556\n",
            "Epoch 16/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3353 - accuracy: 0.8560\n",
            "Epoch 17/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3346 - accuracy: 0.8557\n",
            "Epoch 18/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3341 - accuracy: 0.8563\n",
            "Epoch 19/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3336 - accuracy: 0.8565\n",
            "Epoch 20/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3331 - accuracy: 0.8567\n",
            "Epoch 21/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3325 - accuracy: 0.8571\n",
            "Epoch 22/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3321 - accuracy: 0.8574\n",
            "Epoch 23/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3318 - accuracy: 0.8578\n",
            "Epoch 24/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3313 - accuracy: 0.8575\n",
            "Epoch 25/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3310 - accuracy: 0.8574\n",
            "Epoch 26/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3304 - accuracy: 0.8579\n",
            "Epoch 27/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3303 - accuracy: 0.8578\n",
            "Epoch 28/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3298 - accuracy: 0.8580\n",
            "Epoch 29/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3294 - accuracy: 0.8583\n",
            "Epoch 30/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3290 - accuracy: 0.8583\n",
            "Epoch 31/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3287 - accuracy: 0.8585\n",
            "Epoch 32/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3285 - accuracy: 0.8583\n",
            "Epoch 33/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3282 - accuracy: 0.8589\n",
            "Epoch 34/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3279 - accuracy: 0.8588\n",
            "Epoch 35/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3276 - accuracy: 0.8586\n",
            "Epoch 36/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3274 - accuracy: 0.8588\n",
            "Epoch 37/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3272 - accuracy: 0.8585\n",
            "Epoch 38/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3269 - accuracy: 0.8588\n",
            "Epoch 39/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3267 - accuracy: 0.8588\n",
            "Epoch 40/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3264 - accuracy: 0.8588\n",
            "Epoch 41/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3261 - accuracy: 0.8591\n",
            "Epoch 42/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3260 - accuracy: 0.8593\n",
            "Epoch 43/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3257 - accuracy: 0.8592\n",
            "Epoch 44/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3256 - accuracy: 0.8589\n",
            "Epoch 45/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3255 - accuracy: 0.8595\n",
            "Epoch 46/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3254 - accuracy: 0.8591\n",
            "Epoch 47/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3250 - accuracy: 0.8593\n",
            "Epoch 48/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3249 - accuracy: 0.8592\n",
            "Epoch 49/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3247 - accuracy: 0.8591\n",
            "Epoch 50/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3247 - accuracy: 0.8596\n",
            "Epoch 51/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3247 - accuracy: 0.8592\n",
            "Epoch 52/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3245 - accuracy: 0.8593\n",
            "Epoch 53/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3241 - accuracy: 0.8599\n",
            "Epoch 54/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3240 - accuracy: 0.8601\n",
            "Epoch 55/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3240 - accuracy: 0.8595\n",
            "Epoch 56/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3239 - accuracy: 0.8600\n",
            "Epoch 57/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3238 - accuracy: 0.8598\n",
            "Epoch 58/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3236 - accuracy: 0.8601\n",
            "Epoch 59/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3236 - accuracy: 0.8600\n",
            "Epoch 60/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3235 - accuracy: 0.8597\n",
            "Epoch 61/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3234 - accuracy: 0.8597\n",
            "Epoch 62/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3232 - accuracy: 0.8600\n",
            "Epoch 63/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3230 - accuracy: 0.8601\n",
            "Epoch 64/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3231 - accuracy: 0.8602\n",
            "Epoch 65/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8599\n",
            "Epoch 66/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3229 - accuracy: 0.8603\n",
            "Epoch 67/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3228 - accuracy: 0.8602\n",
            "Epoch 68/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3227 - accuracy: 0.8602\n",
            "Epoch 69/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3226 - accuracy: 0.8600\n",
            "Epoch 70/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3225 - accuracy: 0.8601\n",
            "Epoch 71/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3224 - accuracy: 0.8606\n",
            "Epoch 72/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3222 - accuracy: 0.8605\n",
            "Epoch 73/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3222 - accuracy: 0.8605\n",
            "Epoch 74/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3221 - accuracy: 0.8606\n",
            "Epoch 75/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8606\n",
            "Epoch 76/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8610\n",
            "Epoch 77/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3217 - accuracy: 0.8611\n",
            "Epoch 78/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3219 - accuracy: 0.8610\n",
            "Epoch 79/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3217 - accuracy: 0.8607\n",
            "Epoch 80/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3216 - accuracy: 0.8607\n",
            "Epoch 81/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3213 - accuracy: 0.8611\n",
            "Epoch 82/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3214 - accuracy: 0.8606\n",
            "Epoch 83/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8608\n",
            "Epoch 84/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3211 - accuracy: 0.8607\n",
            "Epoch 85/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3212 - accuracy: 0.8614\n",
            "Epoch 86/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8608\n",
            "Epoch 87/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3211 - accuracy: 0.8611\n",
            "Epoch 88/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8616\n",
            "Epoch 89/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3209 - accuracy: 0.8613\n",
            "Epoch 90/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3210 - accuracy: 0.8609\n",
            "Epoch 91/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3206 - accuracy: 0.8613\n",
            "Epoch 92/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3207 - accuracy: 0.8611\n",
            "Epoch 93/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3206 - accuracy: 0.8614\n",
            "Epoch 94/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3205 - accuracy: 0.8614\n",
            "Epoch 95/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3204 - accuracy: 0.8617\n",
            "Epoch 96/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3203 - accuracy: 0.8619\n",
            "Epoch 97/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3203 - accuracy: 0.8614\n",
            "Epoch 98/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3203 - accuracy: 0.8618\n",
            "Epoch 99/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3202 - accuracy: 0.8619\n",
            "Epoch 100/100\n",
            "112/112 [==============================] - 0s 2ms/step - loss: 0.3201 - accuracy: 0.8614\n",
            "28/28 [==============================] - 0s 1ms/step - loss: 0.3259 - accuracy: 0.8575\n",
            "CPU times: user 2min 1s, sys: 8.58 s, total: 2min 10s\n",
            "Wall time: 2min 4s\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9nwFWPVzNnA7"
      },
      "source": [
        "## Final Results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "D6Xb7NorNxke",
        "outputId": "7b6081dc-e7d5-49ef-9d89-79c653b8a05d"
      },
      "source": [
        "\"Accuracy: %.2f%% (%.2f%%)\" % (results.mean()*100, results.std()*100)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Accuracy: 85.73% (0.12%)'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Qc4tmQcUTX5_"
      },
      "source": [
        "This is the final **result of model accuracy**, after cross-validation scoring.\n",
        "\n",
        "If necessary, we can change the topology or model settings to try to achieve better accuracy values.\n"
      ]
    }
  ]
}