Spam-TP3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "ae42b5cdc085ad0f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.120115Z",
     "start_time": "2024-04-11T15:52:15.893237Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "# Scikit-learn nos ofrece una variedad ampliada de modelos Naive Bayes, para este problema usamos MultinomialNB que es pensado para este tipo de problemas\n",
    "from sklearn.naive_bayes import MultinomialNB   \n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "from sklearn.metrics import roc_auc_score, roc_curve, accuracy_score, precision_score, recall_score, confusion_matrix, ConfusionMatrixDisplay"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fadd4f97a897731c",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "# TP3: Detector de SPAM\n",
    "\n",
    "Uno de los problemas más comunes en la clasificación es la detección de correos electrónicos SPAM. Uno de los primeros modelos utilizados para abordar este problema fue el clasificador de Bayes ingenuo. La detección de SPAM es un problema persistente en el mundo digital, ya que los spammers continúan adaptando sus estrategias para eludir los filtros de correo no deseado. Además del clasificador de Bayes ingenuo, se han desarrollado y utilizado una variedad de técnicas más avanzadas en la detección de SPAM, que incluyen algoritmos de aprendizaje automático, redes neuronales y métodos basados en reglas.\n",
    "\n",
    "En este trabajo práctico, utilizaremos un conjunto de datos que consta de 4601 observaciones de correos electrónicos, de los cuales 2788 son correos legítimos y 1813 son correos SPAM. Dado que el contenido de los correos electrónicos es un tipo de dato no estructurado, es necesario procesarlo de alguna manera. Para este conjunto de datos, ya se ha aplicado un procesamiento típico en el Procesamiento del Lenguaje Natural (NLP), que consiste en contar la frecuencia de palabras observadas en los correos.\n",
    "\n",
    "El procesamiento de lenguaje natural (NLP) desempeña un papel fundamental en la detección de SPAM, ya que permite analizar el contenido de los correos electrónicos y extraer características relevantes para la clasificación. Además de contar la frecuencia de palabras, se pueden utilizar técnicas más sofisticadas, como la extracción de características semánticas y el análisis de sentimientos, para mejorar la precisión de los modelos de detección de SPAM.\n",
    "\n",
    "En este proceso, se cuenta la cantidad de ocurrencias de cada palabra en los diferentes correos.\n",
    "\n",
    "![spam counter](./spam.png)\n",
    "\n",
    "Con el fin de preservar la privacidad de los mensajes, la frecuencia de palabras se encuentra normalizada. El conjunto de datos está compuesto por 54 columnas de atributos que se denominan:\n",
    "\n",
    "- `word_freq_XXXX`: Donde `XXXX` es la palabra o símbolo. Los valores son enteros que van de 0 a 20k.\n",
    "\n",
    "Además, hay una columna adicional llamada `spam`, que es 1 si el correo es SPAM o 0 si no lo es.\n",
    "\n",
    "Los clasificadores de Bayes ingenuos fueron los primeros filtros utilizados por las aplicaciones de correo electrónico, basados en este principio de palabras. La idea es que, partiendo de un dato a priori sobre la probabilidad de que un correo sea SPAM o no, ciertas palabras nos indicarán que la probabilidad a posteriori, dadas esas palabras, es más probable que el correo sea SPAM o no."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "initial_id",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.143226Z",
     "start_time": "2024-04-11T15:52:16.120715Z"
    },
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "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>word_freq_make</th>\n",
       "      <th>word_freq_address</th>\n",
       "      <th>word_freq_all</th>\n",
       "      <th>word_freq_3d</th>\n",
       "      <th>word_freq_our</th>\n",
       "      <th>word_freq_over</th>\n",
       "      <th>word_freq_remove</th>\n",
       "      <th>word_freq_internet</th>\n",
       "      <th>word_freq_order</th>\n",
       "      <th>word_freq_mail</th>\n",
       "      <th>...</th>\n",
       "      <th>char_freq_;</th>\n",
       "      <th>char_freq_(</th>\n",
       "      <th>char_freq_[</th>\n",
       "      <th>char_freq_!</th>\n",
       "      <th>char_freq_$</th>\n",
       "      <th>char_freq_#</th>\n",
       "      <th>capital_run_length_average</th>\n",
       "      <th>capital_run_length_longest</th>\n",
       "      <th>capital_run_length_total</th>\n",
       "      <th>spam</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.778</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.756</td>\n",
       "      <td>61</td>\n",
       "      <td>278</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.21</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.94</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.132</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.372</td>\n",
       "      <td>0.180</td>\n",
       "      <td>0.048</td>\n",
       "      <td>5.114</td>\n",
       "      <td>101</td>\n",
       "      <td>1028</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.23</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.25</td>\n",
       "      <td>...</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.276</td>\n",
       "      <td>0.184</td>\n",
       "      <td>0.010</td>\n",
       "      <td>9.821</td>\n",
       "      <td>485</td>\n",
       "      <td>2259</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.137</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.137</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.537</td>\n",
       "      <td>40</td>\n",
       "      <td>191</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.63</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.537</td>\n",
       "      <td>40</td>\n",
       "      <td>191</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.85</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.85</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.223</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>15</td>\n",
       "      <td>54</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.92</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.64</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.164</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.671</td>\n",
       "      <td>4</td>\n",
       "      <td>112</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.88</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.88</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.206</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2.450</td>\n",
       "      <td>11</td>\n",
       "      <td>49</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.15</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.76</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.271</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.181</td>\n",
       "      <td>0.203</td>\n",
       "      <td>0.022</td>\n",
       "      <td>9.744</td>\n",
       "      <td>445</td>\n",
       "      <td>1257</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.06</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.38</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.244</td>\n",
       "      <td>0.081</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1.729</td>\n",
       "      <td>43</td>\n",
       "      <td>749</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 58 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   word_freq_make  word_freq_address  word_freq_all  word_freq_3d  \\\n",
       "0            0.00               0.64           0.64           0.0   \n",
       "1            0.21               0.28           0.50           0.0   \n",
       "2            0.06               0.00           0.71           0.0   \n",
       "3            0.00               0.00           0.00           0.0   \n",
       "4            0.00               0.00           0.00           0.0   \n",
       "5            0.00               0.00           0.00           0.0   \n",
       "6            0.00               0.00           0.00           0.0   \n",
       "7            0.00               0.00           0.00           0.0   \n",
       "8            0.15               0.00           0.46           0.0   \n",
       "9            0.06               0.12           0.77           0.0   \n",
       "\n",
       "   word_freq_our  word_freq_over  word_freq_remove  word_freq_internet  \\\n",
       "0           0.32            0.00              0.00                0.00   \n",
       "1           0.14            0.28              0.21                0.07   \n",
       "2           1.23            0.19              0.19                0.12   \n",
       "3           0.63            0.00              0.31                0.63   \n",
       "4           0.63            0.00              0.31                0.63   \n",
       "5           1.85            0.00              0.00                1.85   \n",
       "6           1.92            0.00              0.00                0.00   \n",
       "7           1.88            0.00              0.00                1.88   \n",
       "8           0.61            0.00              0.30                0.00   \n",
       "9           0.19            0.32              0.38                0.00   \n",
       "\n",
       "   word_freq_order  word_freq_mail  ...  char_freq_;  char_freq_(  \\\n",
       "0             0.00            0.00  ...         0.00        0.000   \n",
       "1             0.00            0.94  ...         0.00        0.132   \n",
       "2             0.64            0.25  ...         0.01        0.143   \n",
       "3             0.31            0.63  ...         0.00        0.137   \n",
       "4             0.31            0.63  ...         0.00        0.135   \n",
       "5             0.00            0.00  ...         0.00        0.223   \n",
       "6             0.00            0.64  ...         0.00        0.054   \n",
       "7             0.00            0.00  ...         0.00        0.206   \n",
       "8             0.92            0.76  ...         0.00        0.271   \n",
       "9             0.06            0.00  ...         0.04        0.030   \n",
       "\n",
       "   char_freq_[  char_freq_!  char_freq_$  char_freq_#  \\\n",
       "0          0.0        0.778        0.000        0.000   \n",
       "1          0.0        0.372        0.180        0.048   \n",
       "2          0.0        0.276        0.184        0.010   \n",
       "3          0.0        0.137        0.000        0.000   \n",
       "4          0.0        0.135        0.000        0.000   \n",
       "5          0.0        0.000        0.000        0.000   \n",
       "6          0.0        0.164        0.054        0.000   \n",
       "7          0.0        0.000        0.000        0.000   \n",
       "8          0.0        0.181        0.203        0.022   \n",
       "9          0.0        0.244        0.081        0.000   \n",
       "\n",
       "   capital_run_length_average  capital_run_length_longest  \\\n",
       "0                       3.756                          61   \n",
       "1                       5.114                         101   \n",
       "2                       9.821                         485   \n",
       "3                       3.537                          40   \n",
       "4                       3.537                          40   \n",
       "5                       3.000                          15   \n",
       "6                       1.671                           4   \n",
       "7                       2.450                          11   \n",
       "8                       9.744                         445   \n",
       "9                       1.729                          43   \n",
       "\n",
       "   capital_run_length_total  spam  \n",
       "0                       278     1  \n",
       "1                      1028     1  \n",
       "2                      2259     1  \n",
       "3                       191     1  \n",
       "4                       191     1  \n",
       "5                        54     1  \n",
       "6                       112     1  \n",
       "7                        49     1  \n",
       "8                      1257     1  \n",
       "9                       749     1  \n",
       "\n",
       "[10 rows x 58 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = pd.read_csv(\"spambase.csv\") # cargando los datos desde un CSV\n",
    "dataset.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd5762c49a3369dc",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "Para obtener las palábras más usadas podemos hacer un `groupby`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "aa93f67db28da6ff",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.147047Z",
     "start_time": "2024-04-11T15:52:16.143891Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "column_sum = dataset.groupby(by=\"spam\", as_index=False).sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "efab4e7c21dd0461",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "Y despues se pueden combinar las columnas en usando [pd.melt](https://pandas.pydata.org/docs/reference/api/pandas.melt.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "a05390e27958e98b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.150961Z",
     "start_time": "2024-04-11T15:52:16.147743Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "# Obtenemos los atributos y target\n",
    "X = (dataset.drop(columns=\"spam\") * 100).astype(int)\n",
    "#X = dataset2.drop(columns=\"spam\")\n",
    "y = dataset[\"spam\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "686665f4a0ce53e5",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "Se separa el dataset en entrenamiento y evaluación"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "a05dd9a86a0270d1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.155862Z",
     "start_time": "2024-04-11T15:52:16.151805Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test= train_test_split(X, y, test_size = 0.3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4deca5d28cadd43b",
   "metadata": {
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "source": [
    "Escalamos para aplicar en regresión logística"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "712e38ef8afee400",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.162957Z",
     "start_time": "2024-04-11T15:52:16.156768Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.transform(X_test)\n",
    "\n",
    "# Lo transformamos en DataFrames\n",
    "X_train_scaled = pd.DataFrame(X_train_scaled, columns=X.columns)\n",
    "X_test_scaled = pd.DataFrame(X_test_scaled, columns=X.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71beeba0",
   "metadata": {},
   "source": [
    "## Ejercicio 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "76a9e6364b4d74b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-11T15:52:16.168004Z",
     "start_time": "2024-04-11T15:52:16.164460Z"
    },
    "collapsed": false,
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Spam:\n",
      " capital_run_length_total      853233.000\n",
      "capital_run_length_longest    189265.000\n",
      "capital_run_length_average     17258.246\n",
      "word_freq_you                   4105.610\n",
      "word_freq_your                  2502.610\n",
      "word_freq_will                   997.100\n",
      "word_freq_free                   939.790\n",
      "word_freq_our                    931.800\n",
      "char_freq_!                      931.361\n",
      "word_freq_all                    732.080\n",
      "Name: 1, dtype: float64\n",
      "No Spam:\n",
      " capital_run_length_total      450181.000\n",
      "capital_run_length_longest     50782.000\n",
      "capital_run_length_average      6627.915\n",
      "word_freq_you                   3541.710\n",
      "word_freq_george                3527.560\n",
      "word_freq_hp                    2496.580\n",
      "word_freq_will                  1495.270\n",
      "word_freq_your                  1223.100\n",
      "word_freq_hpl                   1204.400\n",
      "word_freq_re                    1159.140\n",
      "Name: 0, dtype: float64\n",
      "\n",
      "Palabras en común en SPAM y No SPAM:\n",
      "{'capital_run_length_average', 'capital_run_length_total', 'word_freq_your', 'word_freq_you', 'word_freq_will', 'capital_run_length_longest'}\n"
     ]
    }
   ],
   "source": [
    "spam_word_frecuencies = column_sum.loc[1].sort_values(ascending=False)\n",
    "no_spam_word_frecuencies = column_sum.loc[0].sort_values(ascending=False)\n",
    "print('Spam:\\n', spam_word_frecuencies.head(10))\n",
    "print('No Spam:\\n', no_spam_word_frecuencies.head(10))\n",
    "\n",
    "common_words = set(spam_word_frecuencies.head(10).index).intersection(set(no_spam_word_frecuencies.head(10).index))\n",
    "print(\"\\nPalabras en común en SPAM y No SPAM:\")\n",
    "print(common_words)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25203039",
   "metadata": {},
   "source": [
    "## Ejercicio 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "32a83687",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test= train_test_split(X, y, test_size = 0.3, random_state=50)\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.transform(X_test)\n",
    "\n",
    "# Lo transformamos en DataFrames\n",
    "X_train_scaled = pd.DataFrame(X_train_scaled, columns=X.columns)\n",
    "X_test_scaled = pd.DataFrame(X_test_scaled, columns=X.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a1142a5",
   "metadata": {},
   "source": [
    "## Ejercicio 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "f679f3eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "naive_bayes_classifier = MultinomialNB()\n",
    "naive_bayes_classifier.fit(X_train, y_train)\n",
    "\n",
    "# Predecir sobre el conjunto de prueba\n",
    "y_pred_naive_bayes_classifier = naive_bayes_classifier.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e327652b",
   "metadata": {},
   "source": [
    "## Ejercicio 4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b6fcf320",
   "metadata": {},
   "outputs": [],
   "source": [
    "logistic_regresion_classifier = LogisticRegression()\n",
    "logistic_regresion_classifier.fit(X_train_scaled, y_train)\n",
    "y_pred_logistic_regresion_classifier = logistic_regresion_classifier.predict(X_test_scaled)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9ec9d07",
   "metadata": {},
   "source": [
    "## Ejercicio 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "976b0302",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Matriz de confusión Bayes Ingenuo:\n",
      "[[727 125]\n",
      " [134 395]]\n",
      "\n",
      "Matriz de confusión Regresion Logistica:\n",
      "[[811  41]\n",
      " [101 428]]\n"
     ]
    }
   ],
   "source": [
    "# Evaluar el rendimiento\n",
    "conf_matrix_naive_bayes_classifier = confusion_matrix(y_test, y_pred_naive_bayes_classifier)\n",
    "conf_matrix_logistic_regresion_classifier = confusion_matrix(y_test, y_pred_logistic_regresion_classifier)\n",
    "\n",
    "\n",
    "print(\"\\nMatriz de confusión Bayes Ingenuo:\")\n",
    "print(conf_matrix_naive_bayes_classifier)\n",
    "print(\"\\nMatriz de confusión Regresion Logistica:\")\n",
    "print(conf_matrix_logistic_regresion_classifier)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "324364aa",
   "metadata": {},
   "source": [
    "## Ejercicio 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "e4feb3d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Exactitud del clasificador de Bayes ingenuo: 0.7596153846153846 , Recuperacion: 0.7466918714555766\n",
      "Exactitud del clasificador de Regresion Logistica: 0.9125799573560768 , Recuperacion: 0.8090737240075614\n"
     ]
    }
   ],
   "source": [
    "# Precisión y recuperación para Bayes ingenuo\n",
    "accuracy_naive_bayes_classifier = precision_score(y_test, y_pred_naive_bayes_classifier)\n",
    "recall_nb = recall_score(y_test, y_pred_naive_bayes_classifier)\n",
    "\n",
    "# Precisión y recuperación para Regresión Logística\n",
    "accuracy_logistic_regresion_classifier = precision_score(y_test, y_pred_logistic_regresion_classifier)\n",
    "recall_lr = recall_score(y_test, y_pred_logistic_regresion_classifier)\n",
    "\n",
    "\n",
    "print(\"Exactitud del clasificador de Bayes ingenuo:\", accuracy_naive_bayes_classifier, \", Recuperacion:\", recall_nb)\n",
    "print(\"Exactitud del clasificador de Regresion Logistica:\", accuracy_logistic_regresion_classifier, \", Recuperacion:\", recall_lr)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0a8dba9",
   "metadata": {},
   "source": [
    "## Ejercicio 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "3d51a063",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC y AUC para Bayes ingenuo\n",
    "y_prob_nb = naive_bayes_classifier.predict_proba(X_test)[:, 1]\n",
    "fpr_nb, tpr_nb, _ = roc_curve(y_test, y_prob_nb)\n",
    "auc_nb = roc_auc_score(y_test, y_prob_nb)\n",
    "\n",
    "# ROC y AUC para Regresión Logística\n",
    "y_prob_lr = logistic_regresion_classifier.predict_proba(X_test_scaled)[:, 1]\n",
    "fpr_lr, tpr_lr, _ = roc_curve(y_test, y_prob_lr)\n",
    "auc_lr = roc_auc_score(y_test, y_prob_lr)\n",
    "\n",
    "plt.figure(figsize=(10, 6))\n",
    "plt.plot(fpr_nb, tpr_nb, label=f'Bayes ingenuo (AUC = {auc_nb:.2f})')\n",
    "plt.plot(fpr_lr, tpr_lr, label=f'Regresión Logística (AUC = {auc_lr:.2f})')\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Curva ROC')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}