pruebaTesis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>Documento</th>\n",
       "      <th>Fecha</th>\n",
       "      <th>Texto</th>\n",
       "      <th>Categoria</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2021-01-02 00:00:00</td>\n",
       "      <td>Primero Emelec y cinco años después el hermano...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2021-01-03 00:00:00</td>\n",
       "      <td>#ElAgroTrabaja por el bienestar de los ecuator...</td>\n",
       "      <td>Social</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2021-01-01 00:00:00</td>\n",
       "      <td>Es nuestro derecho a recibir a tiempo los recu...</td>\n",
       "      <td>Económico</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2020-03-23 00:00:00</td>\n",
       "      <td>Sin distanciamiento Incumpliendo disposiciones...</td>\n",
       "      <td>Social</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>2021-01-03 00:00:00</td>\n",
       "      <td>Guayaquil | La Intendencia de Policía del Guay...</td>\n",
       "      <td>Social</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1335</th>\n",
       "      <td>1336</td>\n",
       "      <td>2021-02-25 00:00:00</td>\n",
       "      <td>Malecón 2000 Music Night Disfruta de música en...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1336</th>\n",
       "      <td>1337</td>\n",
       "      <td>2020-12-02 00:00:00</td>\n",
       "      <td>vivieron emocionantes jornadas deportivas en c...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1337</th>\n",
       "      <td>1338</td>\n",
       "      <td>2021-02-24 00:00:00</td>\n",
       "      <td>Monumento a Spencer les dará la bienvenida al ...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1338</th>\n",
       "      <td>1339</td>\n",
       "      <td>2021-02-25 00:00:00</td>\n",
       "      <td>Renato Ibarra podría salir de Atlas, tras dife...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1339</th>\n",
       "      <td>1340</td>\n",
       "      <td>2021-02-25 00:00:00</td>\n",
       "      <td>La FIFA levantó la sanción a Deportivo Quito y...</td>\n",
       "      <td>no pest</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1340 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Documento                Fecha  \\\n",
       "0            0  2021-01-02 00:00:00   \n",
       "1            1  2021-01-03 00:00:00   \n",
       "2            2  2021-01-01 00:00:00   \n",
       "3            3  2020-03-23 00:00:00   \n",
       "4            4  2021-01-03 00:00:00   \n",
       "...        ...                  ...   \n",
       "1335      1336  2021-02-25 00:00:00   \n",
       "1336      1337  2020-12-02 00:00:00   \n",
       "1337      1338  2021-02-24 00:00:00   \n",
       "1338      1339  2021-02-25 00:00:00   \n",
       "1339      1340  2021-02-25 00:00:00   \n",
       "\n",
       "                                                  Texto  Categoria  \n",
       "0     Primero Emelec y cinco años después el hermano...  no pest    \n",
       "1     #ElAgroTrabaja por el bienestar de los ecuator...     Social  \n",
       "2     Es nuestro derecho a recibir a tiempo los recu...  Económico  \n",
       "3     Sin distanciamiento Incumpliendo disposiciones...     Social  \n",
       "4     Guayaquil | La Intendencia de Policía del Guay...     Social  \n",
       "...                                                 ...        ...  \n",
       "1335  Malecón 2000 Music Night Disfruta de música en...  no pest    \n",
       "1336  vivieron emocionantes jornadas deportivas en c...  no pest    \n",
       "1337  Monumento a Spencer les dará la bienvenida al ...  no pest    \n",
       "1338  Renato Ibarra podría salir de Atlas, tras dife...  no pest    \n",
       "1339  La FIFA levantó la sanción a Deportivo Quito y...  no pest    \n",
       "\n",
       "[1340 rows x 4 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path_file = \"DataSet-pest.xlsx\"\n",
    "col_types = {\"Documento\":str, \"Categoria\":str}\n",
    "df = pd.read_excel(path_file, dtype=col_types)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Documento', 'Fecha', 'Texto', 'Categoria'], dtype=object)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "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>Texto</th>\n",
       "      <th>Categoria</th>\n",
       "      <th>category_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Primero Emelec y cinco años después el hermano...</td>\n",
       "      <td>no pest</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>#ElAgroTrabaja por el bienestar de los ecuator...</td>\n",
       "      <td>Social</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Es nuestro derecho a recibir a tiempo los recu...</td>\n",
       "      <td>Económico</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Sin distanciamiento Incumpliendo disposiciones...</td>\n",
       "      <td>Social</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Guayaquil | La Intendencia de Policía del Guay...</td>\n",
       "      <td>Social</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               Texto  Categoria  category_id\n",
       "0  Primero Emelec y cinco años después el hermano...  no pest              0\n",
       "1  #ElAgroTrabaja por el bienestar de los ecuator...     Social            1\n",
       "2  Es nuestro derecho a recibir a tiempo los recu...  Económico            2\n",
       "3  Sin distanciamiento Incumpliendo disposiciones...     Social            1\n",
       "4  Guayaquil | La Intendencia de Policía del Guay...     Social            1"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from io import StringIO\n",
    "col = ['Texto', 'Categoria']\n",
    "df = df[col]\n",
    "df = df[pd.notnull(df['Categoria'])]\n",
    "\n",
    "df.columns = ['Texto', 'Categoria']\n",
    "\n",
    "df['category_id'] = df['Categoria'].factorize()[0]\n",
    "category_id_df = df[['Categoria', 'category_id']].drop_duplicates().sort_values('category_id')\n",
    "category_to_id = dict(category_id_df.values)\n",
    "id_to_category = dict(category_id_df[['category_id','Categoria']].values)\n",
    "\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "fig = plt.figure(figsize=(8,6))\n",
    "df.groupby('Categoria').Texto.count().plot.bar(ylim=0)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       no pest  \n",
       "1          Social\n",
       "2       Económico\n",
       "3          Social\n",
       "4          Social\n",
       "5          Social\n",
       "6          Social\n",
       "7          Social\n",
       "8          Social\n",
       "9          Social\n",
       "10         Social\n",
       "11    Tecnológico\n",
       "12    Tecnológico\n",
       "13         Social\n",
       "14         Social\n",
       "15         Social\n",
       "16         Social\n",
       "17         Social\n",
       "18         Social\n",
       "19         Social\n",
       "Name: Categoria, dtype: object"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Categoria[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1340, 1345)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from nltk.corpus import stopwords \n",
    "tfidf = TfidfVectorizer (sublinear_tf = True, min_df = 5,norm = 'l2', encoding = 'latin-1', ngram_range = (1, 2), stop_words = 'english')\n",
    "count_vect = CountVectorizer()\n",
    "# features = tfidf.fit_transform(df['Categoria']) .toarray()\n",
    "features = tfidf.fit_transform(df['Texto'].values.astype('U')) .toarray()\n",
    "labels = df.Categoria\n",
    "features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1340, 6525)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "count_vect = CountVectorizer()\n",
    "X_train_counts = count_vect.fit_transform(df['Texto'].values.astype('U')) .toarray()\n",
    "X_train_counts.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1340, 6525)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "tf_transformer = TfidfTransformer(use_idf=False).fit(X_train_counts)\n",
    "X_train_tf = tf_transformer.transform(X_train_counts)\n",
    "X_train_tf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import MultinomialNB\n",
    "\n",
    "clf = MultinomialNB().fit(features, df.Categoria)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 'Económico':\n",
      " .La mayoría de los unigramas correlacionados:\n",
      ". elgobiernodetodos\n",
      ".último\n",
      " .Bigrams más correlacionados:\n",
      ". emergencia sanitaria\n",
      ".empleo para\n",
      "# 'Político':\n",
      " .La mayoría de los unigramas correlacionados:\n",
      ". elgobiernodetodos\n",
      ".último\n",
      " .Bigrams más correlacionados:\n",
      ". emergencia sanitaria\n",
      ".empleo para\n",
      "# 'Social':\n",
      " .La mayoría de los unigramas correlacionados:\n",
      ". elgobiernodetodos\n",
      ".último\n",
      " .Bigrams más correlacionados:\n",
      ". emergencia sanitaria\n",
      ".empleo para\n",
      "# 'Tecnológico':\n",
      " .La mayoría de los unigramas correlacionados:\n",
      ". elgobiernodetodos\n",
      ".último\n",
      " .Bigrams más correlacionados:\n",
      ". emergencia sanitaria\n",
      ".empleo para\n",
      "# 'no pest  ':\n",
      " .La mayoría de los unigramas correlacionados:\n",
      ". elgobiernodetodos\n",
      ".último\n",
      " .Bigrams más correlacionados:\n",
      ". emergencia sanitaria\n",
      ".empleo para\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_selection import chi2\n",
    "import numpy as np\n",
    "\n",
    "N = 2\n",
    "for Categoria, category_id in sorted(category_to_id.items()):\n",
    "    features_chi2 = chi2(features, labels == category_id)\n",
    "    indices = np.argsort(features_chi2[0])\n",
    "    feature_names = np.array(tfidf.get_feature_names()) [indices]\n",
    "    unigrams = [v for v in feature_names if len(v.split(' ')) == 1]\n",
    "    bigrams = [v for v in feature_names if len(v.split(' ')) == 2]\n",
    "    print(\"# '{}':\".format(Categoria))\n",
    "    print(\" .La mayoría de los unigramas correlacionados:\\n. {}\".format('\\n.'.join(unigrams[-N:])))\n",
    "    print(\" .Bigrams más correlacionados:\\n. {}\".format('\\n.'.join(bigrams[-N:])))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(df['Texto'], df['Categoria'],\n",
    "random_state = 0)\n",
    "count_vect = CountVectorizer()\n",
    "X_train_counts = count_vect.fit_transform(X_train)\n",
    "tfidf_transformer = TfidfTransformer()\n",
    "X_train_tfidf = tfidf_transformer.fit_transform(X_train_counts)\n",
    "\n",
    "clf = MultinomialNB().fit(X_train_tfidf, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Social']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Gastronomía y deporte se fusionan en homenaje a las fiestas julianas \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Económico']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Banco Central: La economía ecuatoriana se contrajo 12,4 % en el segundo trimestre de 2020\"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Político']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Candidato Andrés Arauz lidera intención de voto en Ecuador \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Tecnológico']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" comercio electrónico \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Social']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" La ONU y la CIDH piden respuestas del Estado por masacre en cárceles; 33 colectivos hacen siete pedidos \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Económico']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Inversión extranjera directa cae un 42 % durante la pandemia, son $107 millones menos que entraron a Ecuador \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Económico']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Sector turístico impulsa plataforma digital para incentivar viajes internos \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Político']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([\" Distrito de Salud entrega reconocimiento como Establecimientos 100% Libres de Humo de Tabaco a las instituciones públicas: Jefatura Política y Consejo de Protección de Derechos de Nobol, y  Jefatura política, Dinapen y Consejo de Segurida Ciudadana del cantón Daule. \"])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x560 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "models = [\n",
    "    RandomForestClassifier(n_estimators=200, max_depth=3, \n",
    "random_state=0),\n",
    "         LinearSVC(),\n",
    "         MultinomialNB(),\n",
    "         LogisticRegression(random_state=0),\n",
    "    SVC(gamma=2,C=1)\n",
    "]\n",
    "CV = 5\n",
    "cv_df = pd.DataFrame(index=range(CV * len(models)))\n",
    "entries = []\n",
    "for model in models:\n",
    "    model_name =model.__class__.__name__\n",
    "    accuracies = cross_val_score(model, features, labels, scoring='accuracy', cv=CV)\n",
    "    for fold_idx, accuracy in enumerate (accuracies):\n",
    "        entries.append((model_name, fold_idx, accuracy))\n",
    "    cv_df = pd.DataFrame(entries, columns=['model_name','fold_idx','accuracy'])\n",
    "    \n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "w = 8\n",
    "h = 7\n",
    "d = 80\n",
    "\n",
    "plt.figure(figsize=(w,h), dpi=d)\n",
    "sns.boxplot(x='model_name', y='accuracy', data=cv_df, width=0.9, fliersize=4, linewidth=2.5, whis= 4.6)\n",
    "sns.stripplot(x='model_name', y='accuracy', data=cv_df, size=8, jitter=True, edgecolor=\"gray\", linewidth=2)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "model_name\n",
       "LinearSVC                 0.608209\n",
       "LogisticRegression        0.591791\n",
       "MultinomialNB             0.602985\n",
       "RandomForestClassifier    0.446269\n",
       "SVC                       0.551493\n",
       "Name: accuracy, dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_df.groupby('model_name').accuracy.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Matriz de confusión\n",
    "model = LinearSVC()\n",
    "\n",
    "X_train, X_test, y_train, y_test, indices_train, indices_test = train_test_split(features, labels, df.index, test_size=0.33,\n",
    "random_state=0)\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "conf_mat = confusion_matrix(y_test, y_pred)\n",
    "fig, ax = plt.subplots(figsize=(10,10))\n",
    "sns.heatmap(conf_mat, annot=True, fmt='d',\n",
    "           xticklabels=category_id_df.Categoria.values,\n",
    "yticklabels=category_id_df.Categoria.values)\n",
    "plt.ylabel('Actual')\n",
    "plt.xlabel('Predicted')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "   no pest         0.69      0.68      0.68       102\n",
      "      Social       0.76      0.71      0.73       123\n",
      "   Económico       0.66      0.83      0.73       146\n",
      " Tecnológico       0.57      0.42      0.48        38\n",
      "    Político       0.71      0.35      0.47        34\n",
      "\n",
      "    accuracy                           0.69       443\n",
      "   macro avg       0.68      0.60      0.62       443\n",
      "weighted avg       0.69      0.69      0.68       443\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(metrics.classification_report(y_test, y_pred, target_names=df['Categoria'].unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(443, 1345)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}