P06_01_Categorisez_automatiquement_des_questions_Nettoyage.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1e420fec",
   "metadata": {
    "id": "1e420fec"
   },
   "source": [
    "# Categorisez automatiquement des questions - Nettoyage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "su1aHDH8FVTd",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "su1aHDH8FVTd",
    "outputId": "c2478f56-9c4e-4492-a40b-e77271769e46"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipping\n"
     ]
    }
   ],
   "source": [
    "%%script echo skipping\n",
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d122f713",
   "metadata": {
    "id": "d122f713"
   },
   "source": [
    "# SOMMAIRE\n",
    "* [PREPARATION](#PREPARATION)\n",
    "* [CHARGEMENT](#CHARGEMENT)\n",
    "* [CARACTERISTIQUES](#CARACTERISTIQUES)\n",
    "* [NETTOYAGE](#NETTOYAGE)\n",
    "    * [Tokenisation](#Tokenisation)\n",
    "    * [Nettoyage des stopwords](#Nettoyage-des-stopwords)\n",
    "    * [Lemmatisation](#Lemmatisation)\n",
    "* [SAUVEGARDE](#SAUVEGARDE)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71f62611",
   "metadata": {
    "id": "71f62611"
   },
   "source": [
    "# PREPARATION"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "753aaabc",
   "metadata": {
    "id": "753aaabc"
   },
   "source": [
    "## Import des librairies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e5151aa9",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e5151aa9",
    "outputId": "fde43a2f-58e1-4603-a460-99e99871807d"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to /Users/hui-guan/nltk_data...\n",
      "[nltk_data]   Unzipping tokenizers/punkt.zip.\n",
      "[nltk_data] Downloading package stopwords to /Users/hui-\n",
      "[nltk_data]     guan/nltk_data...\n",
      "[nltk_data]   Unzipping corpora/stopwords.zip.\n",
      "[nltk_data] Downloading package averaged_perceptron_tagger to\n",
      "[nltk_data]     /Users/hui-guan/nltk_data...\n",
      "[nltk_data]   Unzipping taggers/averaged_perceptron_tagger.zip.\n",
      "[nltk_data] Downloading package wordnet to /Users/hui-\n",
      "[nltk_data]     guan/nltk_data...\n",
      "[nltk_data]   Unzipping corpora/wordnet.zip.\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "import matplotlib as mplt\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "import joblib\n",
    "\n",
    "import re\n",
    "\n",
    "import nltk\n",
    "nltk.download('punkt')\n",
    "nltk.download('stopwords')\n",
    "nltk.download('averaged_perceptron_tagger')\n",
    "nltk.download('wordnet')\n",
    "from nltk.corpus import stopwords, wordnet\n",
    "from nltk.stem import WordNetLemmatizer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fdd5477b",
   "metadata": {
    "id": "fdd5477b"
   },
   "source": [
    "## Paramétrage du notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c07193e3",
   "metadata": {
    "id": "c07193e3"
   },
   "outputs": [],
   "source": [
    "from matplotlib import font_manager\n",
    "\n",
    "font_dirs = ['/Library/Fonts/']\n",
    "font_files = font_manager.findSystemFonts(fontpaths=font_dirs)\n",
    "\n",
    "for font_file in font_files:\n",
    "    font_manager.fontManager.addfont(font_file)\n",
    "\n",
    "# set font\n",
    "plt.rcParams['font.family'] = 'Comic Sans'\n",
    "plt.rcParams['font.sans-serif'] = 'Franklin Gothic Book'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ef868628",
   "metadata": {
    "id": "ef868628"
   },
   "outputs": [],
   "source": [
    "sns.set(rc={\n",
    " 'axes.axisbelow': False,\n",
    " 'axes.edgecolor': 'lightgrey',\n",
    " 'axes.facecolor': 'None',\n",
    " 'axes.grid': False,\n",
    " 'axes.labelcolor': 'dimgrey',\n",
    " 'axes.spines.right': False,\n",
    " 'axes.spines.top': False,\n",
    " 'figure.facecolor': 'white',\n",
    " 'lines.solid_capstyle': 'round',\n",
    " 'patch.edgecolor': 'w',\n",
    " 'patch.force_edgecolor': True,\n",
    " 'text.color': 'dimgrey',\n",
    " 'xtick.bottom': False,\n",
    " 'xtick.color': 'dimgrey',\n",
    " 'xtick.direction': 'out',\n",
    " 'xtick.top': False,\n",
    " 'ytick.color': 'dimgrey',\n",
    " 'ytick.direction': 'out',\n",
    " 'ytick.left': False,\n",
    " 'ytick.right': False})\n",
    "\n",
    "#mplt.font_manager._rebuild()\n",
    "#mplt.font_manager.findfont('Franklin Gothic Book',rebuild_if_missing=True)\n",
    "\n",
    "\n",
    "pd.options.display.max_colwidth=120\n",
    "pd.options.display.min_rows=100"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3961885b",
   "metadata": {
    "id": "3961885b"
   },
   "source": [
    "## Declaration des fonctions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "86cf8522",
   "metadata": {
    "id": "86cf8522"
   },
   "outputs": [],
   "source": [
    "#%run /Users/hui-guan/Documents/Developpement/Jupyter/Lib/outilsEDA.py\n",
    "\n",
    "def profil(df):\n",
    "    \"\"\"\n",
    "    Fonction qui retourne le profil des données d'un dataframe (moyenne, ecart type, min, max, valeurs manquantes, valeurs uniques)\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    result : DataFrame\n",
    "        un DataFrame de profil, contenant un résumé statistique des données\n",
    "    \"\"\"\n",
    "    \n",
    "    dfValeursManquantes=pd.concat([df.isna().sum(),100*df.isna().sum()/len(df)],axis=1,keys=[\"missing_values\",\"%_missing_values\"])\n",
    "    dfTypes=pd.DataFrame({'type':df.dtypes})\n",
    "    result=df.describe(include='all',datetime_is_numeric=True).T\n",
    "    result=result.merge(dfTypes,left_index=True,right_index=True)\n",
    "    result=result.merge(dfValeursManquantes,left_index=True,right_index=True)\n",
    "    return result\n",
    "def summarizeFrequenciesDf(df,threshold):\n",
    "    \"\"\"\n",
    "    Fonction mode qui agrege les variables qui sont sous un seuil 'threshold'(pour les très gros dataframe qui ont du mal à plotter)\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "    threshold: Int\n",
    "        seuil au dessous duquel le comptage des observations sont agrégées dans la catégorie 'autres'\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    result : DataFrame\n",
    "        un DataFrame contenant le comptage des observations au dessus du seuil et tout le reste dans la catégorie 'Autres'\n",
    "    \"\"\"\n",
    "\n",
    "    dfFrequencies=df.value_counts()\n",
    "    dfNormalizedFrequencies=df.value_counts(normalize=True) # frequence normalisée des valeurs de chaque variable\n",
    "    mask = dfNormalizedFrequencies>threshold # masque sur les fréquences > seuil\n",
    "    dfTail= dfFrequencies.loc[~mask].sum()\n",
    "    dfFrequencies=dfFrequencies.loc[mask] # filtre sur les fréquences < seuil\n",
    "    dfFrequencies['Autres']=dfTail # ajout des fréquences > seuil agregées dans la categorie 'Autre'\n",
    "    return pd.DataFrame(dfFrequencies)\n",
    "def plotCategoricalDistributions(plt,df,nrows,ncols,subfigwidth,subfigheight):\n",
    "    \"\"\"\n",
    "    Fonction affiche les distributions des variables categorielles d'un DataFrame\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    plt: plot\n",
    "        un objet pyplot qui affiche les graphiques\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "    nrows: Int\n",
    "        nombre de lignes des sous-figures\n",
    "    ncols: Int\n",
    "        nombre de colonnes des sous-figures\n",
    "    subfigwidth: Int\n",
    "        largeur des sous-figures\n",
    "    subfigheight: Int\n",
    "        hauteur des sous-figures\n",
    "\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    Pas de retours, c'est une procédure d'affichage\n",
    "\n",
    "    \"\"\"\n",
    "    dfCategorical=df.select_dtypes(include=['object','string', 'category'])\n",
    "    \n",
    "    fig, axes = plt.subplots(nrows=nrows, ncols=ncols,figsize=(subfigwidth,subfigheight))\n",
    "\n",
    "    # Titre\n",
    "    fig.suptitle(\"Distribution Variables Categorielles\", fontsize=20)\n",
    "    fig.subplots_adjust(top=1.5)\n",
    "    \n",
    "    # axes est un tableau à 2 dimensions: lignes, colonnes\n",
    "    # pour parcourir les subplot, il faut réduire axes à une dimension avec flatten\n",
    "    axes=axes.flatten()\n",
    "    \n",
    "    # espace vertical entre les rangées de sous graphiques\n",
    "    plt.tight_layout(pad=3)\n",
    "    \n",
    "    i=0\n",
    "    for x in dfCategorical.columns:\n",
    "        # passe les labels sur l'axe des x en arriere plan\n",
    "        axes[i].xaxis.label.set_visible(False)\n",
    "        \n",
    "        # On classe dans la categorie \"Autres\" les variables sous le seuil de 0,5% de frequence d'observation,\n",
    "        dfCurrent=summarizeFrequenciesDf(dfCategorical[x],0.005)\n",
    "    \n",
    "        axes[i].bar(dfCurrent[x].index,dfCurrent[x],color='#ee6352')\n",
    "        axes[i].tick_params(axis='x', rotation=-90)\n",
    "        axes[i].set_title(x,color='#3fa7d6')\n",
    "    \n",
    "        i=i+1\n",
    "    plt.show()\n",
    "        \n",
    "def plotQuantitativeDistributions(plt,df,nrows,ncols,subfigwidth,subfigheight):\n",
    "    \"\"\"\n",
    "    Fonction affiche les distributions des variables quantitatives d'un DataFrame\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    plt: plot\n",
    "        un objet pyplot qui affiche les graphiques\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "    nrows: Int\n",
    "        nombre de lignes des sous-figures\n",
    "    ncols: Int\n",
    "        nombre de colonnes des sous-figures\n",
    "    subfigwidth: Int\n",
    "        largeur des sous-figures\n",
    "    subfigheight: Int\n",
    "        hauteur des sous-figures\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    Pas de retours, c'est une procédure d'affichage\n",
    "\n",
    "    \"\"\"\n",
    "    dfQuantitative=df.select_dtypes(exclude=['object','string', 'category'])\n",
    "    dfQSkew=dfQuantitative.skew(skipna = True).to_frame('SKEW')\n",
    "    dfQColumnsSorted=dfQSkew.sort_values(by='SKEW',ascending=False).T.columns #tri des colonnes par valeur de skew\n",
    "\n",
    "    fig, axes = plt.subplots(nrows=nrows, ncols=ncols,figsize=(subfigwidth,subfigheight))\n",
    "    \n",
    "    # Titre\n",
    "    fig.suptitle(\"Distribution Variables Quantitatives\", fontsize=20)\n",
    "    fig.subplots_adjust(top=1.5)\n",
    "    \n",
    "    # axes est un tableau à 2 dimensions: lignes, colonnes\n",
    "    # pour parcourir les subplot, il faut réduire axes à une dimension avec flatten\n",
    "    axes=axes.flatten()\n",
    "    \n",
    "    # espace vertical entre les rangées de sous graphiques\n",
    "    plt.tight_layout(pad=3)\n",
    "    \n",
    "    dfQuantitative\n",
    "    i=0\n",
    "    for x in dfQColumnsSorted:\n",
    "        #dfCurrent=dfNumerical[x].sample(1000,replace=True).sort_index() #optimisation pour les large datasets\n",
    "        dfCurrent=dfQuantitative[x].dropna(axis=0)\n",
    "        # passe les labels sur l'axe des x en arriere plan\n",
    "        axes[i].xaxis.label.set_visible(False)    \n",
    "        \n",
    "        axes[i].hist(dfCurrent,color='#3fa7d6')\n",
    "        axes[i].tick_params(axis='x', rotation=-90)\n",
    "        axes[i].set_title(x,color='#ee6352')\n",
    "    \n",
    "        i=i+1\n",
    "    plt.show()\n",
    "def plotCorrelationHeatmap(plt,df):\n",
    "    \"\"\"\n",
    "    Fonction affiche la heatmap des correlation d'un DataFrame\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    plt: plot\n",
    "        un objet pyplot qui affiche les graphiques\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    Pas de retours, c'est une procédure d'affichage\n",
    "\n",
    "    \"\"\"\n",
    "    dfCorr=df.corr()\n",
    "    plt.rcParams[\"figure.figsize\"] = (10,10)\n",
    "    plt.matshow(dfCorr,cmap='coolwarm',interpolation='nearest')\n",
    "    plt.xticks(range(len(dfCorr.columns)), dfCorr.columns, size='small',rotation='vertical');\n",
    "    plt.yticks(range(len(dfCorr.columns)), dfCorr.columns, size='small');\n",
    "    plt.show()\n",
    "    \n",
    "def plotBoxes(plt,df,nrows,ncols):\n",
    "    \"\"\"\n",
    "    Fonction affiche les boites à moustache des variables quantitatives d'un DataFrame\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    plt: plot\n",
    "        un objet pyplot qui affiche les graphiques\n",
    "    df : DataFrame\n",
    "        un DataFrame de données\n",
    "    nrows: Int\n",
    "        nombre de lignes des sous-figures\n",
    "    ncols: Int\n",
    "        nombre de colonnes des sous-figures\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    Pas de retours, c'est une procédure d'affichage\n",
    "    \"\"\"\n",
    "    dfQuantitative=df.select_dtypes(exclude=['object','string', 'category'])\n",
    "\n",
    "    fig, axes = plt.subplots(nrows=nrows, ncols=ncols,figsize=(14,20))\n",
    "    \n",
    "    # espace vertical entre les rangées de sous graphiques\n",
    "    plt.tight_layout(pad=3)\n",
    "\n",
    "\n",
    "    # axes est un tableau à 2 dimensions: lignes, colonnes\n",
    "    # pour parcourir les subplot, il faut réduire axes à une dimension avec flatten\n",
    "    axes=axes.flatten()\n",
    "    \n",
    "    i=0\n",
    "    for x in dfQuantitative.columns:\n",
    "            axes[i].boxplot(dfQuantitative[x].dropna())\n",
    "            axes[i].set_title(x,color='#3fa7d6')\n",
    "            i+=1\n",
    "    plt.show()\n",
    "\n",
    "def tokenize(text):\n",
    "    tokens=re.findall(r'\\w{2,}', text)\n",
    "    return tokens\n",
    "# On remarque que la lemmatisation de base n'est pas tout à fait correcte: \n",
    "# Parce que \"are\" n'est pas converti en \"to be\" et \"hanging\" n'est pas converti en \"hang\" comme attendu. \n",
    "# Cela peut être corrigé si nous fournissons la balise \"part of speech\" correcte (balise POS) comme deuxième argument de lemmatize()\n",
    "def get_wordnet_pos(word):\n",
    "    \"\"\"Map POS tag to first character lemmatize() accepts\"\"\"\n",
    "    tag = nltk.pos_tag([word])[0][1][0].upper()\n",
    "    tag_dict = {\"J\": wordnet.ADJ,\n",
    "                \"N\": wordnet.NOUN,\n",
    "                \"V\": wordnet.VERB,\n",
    "                \"R\": wordnet.ADV}\n",
    "\n",
    "    return tag_dict.get(tag, wordnet.NOUN)\n",
    "\n",
    "def lemmatize(tokens):\n",
    "    lemmatizer = WordNetLemmatizer()\n",
    "    lemmas = [lemmatizer.lemmatize(token,get_wordnet_pos(token)) for token in tokens]\n",
    "    return lemmas"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c55a160a",
   "metadata": {
    "id": "c55a160a"
   },
   "source": [
    "# CHARGEMENT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bcdc35e8",
   "metadata": {
    "id": "bcdc35e8"
   },
   "outputs": [],
   "source": [
    "# Calculate the time elapsed in seconds\n",
    "start = time.time()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e5608c46",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "e5608c46",
    "outputId": "efce8648-97dc-4bbd-fc23-a359f5b52708",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50000 entries, 0 to 49999\n",
      "Data columns (total 23 columns):\n",
      " #   Column                 Non-Null Count  Dtype  \n",
      "---  ------                 --------------  -----  \n",
      " 0   Id                     50000 non-null  int64  \n",
      " 1   PostTypeId             50000 non-null  int64  \n",
      " 2   AcceptedAnswerId       29765 non-null  float64\n",
      " 3   ParentId               0 non-null      float64\n",
      " 4   CreationDate           50000 non-null  object \n",
      " 5   DeletionDate           0 non-null      float64\n",
      " 6   Score                  50000 non-null  int64  \n",
      " 7   ViewCount              50000 non-null  int64  \n",
      " 8   Body                   50000 non-null  object \n",
      " 9   OwnerUserId            49047 non-null  float64\n",
      " 10  OwnerDisplayName       2786 non-null   object \n",
      " 11  LastEditorUserId       27684 non-null  float64\n",
      " 12  LastEditorDisplayName  1531 non-null   object \n",
      " 13  LastEditDate           28133 non-null  object \n",
      " 14  LastActivityDate       50000 non-null  object \n",
      " 15  Title                  50000 non-null  object \n",
      " 16  Tags                   50000 non-null  object \n",
      " 17  AnswerCount            50000 non-null  int64  \n",
      " 18  CommentCount           50000 non-null  int64  \n",
      " 19  FavoriteCount          15172 non-null  float64\n",
      " 20  ClosedDate             2267 non-null   object \n",
      " 21  CommunityOwnedDate     177 non-null    object \n",
      " 22  ContentLicense         50000 non-null  object \n",
      "dtypes: float64(6), int64(6), object(11)\n",
      "memory usage: 8.8+ MB\n",
      "CPU times: user 1.11 s, sys: 175 ms, total: 1.29 s\n",
      "Wall time: 1.3 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "#df=pd.read_csv(\"drive/MyDrive/Colab Notebooks/Datasets/stackoverfllow_qa.csv\")\n",
    "df=pd.read_csv(\"../Dataset/stackoverfllow_qa.csv\")\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9f7162d2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 904
    },
    "id": "9f7162d2",
    "outputId": "35b00bf0-c57c-4b8b-b5c7-46d8bcf8c1fd"
   },
   "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>Id</th>\n",
       "      <th>PostTypeId</th>\n",
       "      <th>AcceptedAnswerId</th>\n",
       "      <th>ParentId</th>\n",
       "      <th>CreationDate</th>\n",
       "      <th>DeletionDate</th>\n",
       "      <th>Score</th>\n",
       "      <th>ViewCount</th>\n",
       "      <th>Body</th>\n",
       "      <th>OwnerUserId</th>\n",
       "      <th>OwnerDisplayName</th>\n",
       "      <th>LastEditorUserId</th>\n",
       "      <th>LastEditorDisplayName</th>\n",
       "      <th>LastEditDate</th>\n",
       "      <th>LastActivityDate</th>\n",
       "      <th>Title</th>\n",
       "      <th>Tags</th>\n",
       "      <th>AnswerCount</th>\n",
       "      <th>CommentCount</th>\n",
       "      <th>FavoriteCount</th>\n",
       "      <th>ClosedDate</th>\n",
       "      <th>CommunityOwnedDate</th>\n",
       "      <th>ContentLicense</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11116</th>\n",
       "      <td>52130806</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2018-09-01 17:58:55</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-3</td>\n",
       "      <td>71</td>\n",
       "      <td>&lt;p&gt;I want to create dynamic number of columns based on some condition given in R:&lt;/p&gt;\\n\\n&lt;p&gt;Say,I've a dataframe cal...</td>\n",
       "      <td>9999850.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6326565.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2018-09-02 14:44:07</td>\n",
       "      <td>2018-09-02 14:44:07</td>\n",
       "      <td>How to create specific number of columns dynamically based on a condition given a column in R</td>\n",
       "      <td>&lt;r&gt;&lt;dataframe&gt;</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC BY-SA 4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44054</th>\n",
       "      <td>62820834</td>\n",
       "      <td>1</td>\n",
       "      <td>63740881.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020-07-09 17:53:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1968</td>\n",
       "      <td>&lt;p&gt;I am running into below error during terraform apply.&lt;/p&gt;\\n&lt;pre&gt;&lt;code&gt;Error: timeout while waiting for plugin to ...</td>\n",
       "      <td>12368716.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1839482.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020-07-09 18:36:40</td>\n",
       "      <td>2020-09-04 12:05:46</td>\n",
       "      <td>Terraform apply throwing Error \"timeout while waiting for plugin to start\"</td>\n",
       "      <td>&lt;kubernetes&gt;&lt;terraform&gt;&lt;terraform-provider-aws&gt;&lt;iaas&gt;&lt;terraform0.12+&gt;</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC BY-SA 4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34879</th>\n",
       "      <td>3403629</td>\n",
       "      <td>1</td>\n",
       "      <td>3403744.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2010-08-04 07:58:08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>912</td>\n",
       "      <td>&lt;p&gt;I am generating an automatically generated periodical report with Sweave. In order to create a nice header I use ...</td>\n",
       "      <td>366256.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3885376.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-10-29 19:56:59</td>\n",
       "      <td>2017-10-29 19:56:59</td>\n",
       "      <td>How to use Sweave (R) code inside a fancyhdr?</td>\n",
       "      <td>&lt;r&gt;&lt;latex&gt;&lt;sweave&gt;</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC BY-SA 3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43463</th>\n",
       "      <td>54154888</td>\n",
       "      <td>1</td>\n",
       "      <td>54155223.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2019-01-11 22:28:49</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2</td>\n",
       "      <td>42</td>\n",
       "      <td>&lt;p&gt;Whenever I run this I got something like ' 196875307' as the total, could \\nsomeone tell me whats wrong with it.H...</td>\n",
       "      <td>10850719.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10850719.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2019-01-11 22:47:18</td>\n",
       "      <td>2019-01-11 23:09:30</td>\n",
       "      <td>Can someone explain me whats wrong with finding total part in this code</td>\n",
       "      <td>&lt;c&gt;</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC BY-SA 4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27023</th>\n",
       "      <td>3385820</td>\n",
       "      <td>1</td>\n",
       "      <td>3385844.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2010-08-02 07:31:12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>15188</td>\n",
       "      <td>&lt;p&gt;HI,&lt;/p&gt;\\n\\n&lt;p&gt;I'm having a JSON parsed return object set.&lt;/p&gt;\\n\\n&lt;pre&gt;&lt;code&gt;{\\n  \"word\":[\\n      \"offered\",\\n    ...</td>\n",
       "      <td>403255.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21234.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2010-08-02 07:31:47</td>\n",
       "      <td>2012-08-03 08:43:30</td>\n",
       "      <td>JSON - Find the length of objects</td>\n",
       "      <td>&lt;javascript&gt;&lt;json&gt;</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CC BY-SA 2.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id  PostTypeId  AcceptedAnswerId  ParentId         CreationDate  \\\n",
       "11116  52130806           1               NaN       NaN  2018-09-01 17:58:55   \n",
       "44054  62820834           1        63740881.0       NaN  2020-07-09 17:53:08   \n",
       "34879   3403629           1         3403744.0       NaN  2010-08-04 07:58:08   \n",
       "43463  54154888           1        54155223.0       NaN  2019-01-11 22:28:49   \n",
       "27023   3385820           1         3385844.0       NaN  2010-08-02 07:31:12   \n",
       "\n",
       "       DeletionDate  Score  ViewCount  \\\n",
       "11116           NaN     -3         71   \n",
       "44054           NaN      2       1968   \n",
       "34879           NaN      3        912   \n",
       "43463           NaN     -2         42   \n",
       "27023           NaN      0      15188   \n",
       "\n",
       "                                                                                                                          Body  \\\n",
       "11116  <p>I want to create dynamic number of columns based on some condition given in R:</p>\\n\\n<p>Say,I've a dataframe cal...   \n",
       "44054  <p>I am running into below error during terraform apply.</p>\\n<pre><code>Error: timeout while waiting for plugin to ...   \n",
       "34879  <p>I am generating an automatically generated periodical report with Sweave. In order to create a nice header I use ...   \n",
       "43463  <p>Whenever I run this I got something like ' 196875307' as the total, could \\nsomeone tell me whats wrong with it.H...   \n",
       "27023  <p>HI,</p>\\n\\n<p>I'm having a JSON parsed return object set.</p>\\n\\n<pre><code>{\\n  \"word\":[\\n      \"offered\",\\n    ...   \n",
       "\n",
       "       OwnerUserId OwnerDisplayName  LastEditorUserId LastEditorDisplayName  \\\n",
       "11116    9999850.0              NaN         6326565.0                   NaN   \n",
       "44054   12368716.0              NaN         1839482.0                   NaN   \n",
       "34879     366256.0              NaN         3885376.0                   NaN   \n",
       "43463   10850719.0              NaN        10850719.0                   NaN   \n",
       "27023     403255.0              NaN           21234.0                   NaN   \n",
       "\n",
       "              LastEditDate     LastActivityDate  \\\n",
       "11116  2018-09-02 14:44:07  2018-09-02 14:44:07   \n",
       "44054  2020-07-09 18:36:40  2020-09-04 12:05:46   \n",
       "34879  2017-10-29 19:56:59  2017-10-29 19:56:59   \n",
       "43463  2019-01-11 22:47:18  2019-01-11 23:09:30   \n",
       "27023  2010-08-02 07:31:47  2012-08-03 08:43:30   \n",
       "\n",
       "                                                                                               Title  \\\n",
       "11116  How to create specific number of columns dynamically based on a condition given a column in R   \n",
       "44054                     Terraform apply throwing Error \"timeout while waiting for plugin to start\"   \n",
       "34879                                                  How to use Sweave (R) code inside a fancyhdr?   \n",
       "43463                        Can someone explain me whats wrong with finding total part in this code   \n",
       "27023                                                              JSON - Find the length of objects   \n",
       "\n",
       "                                                                        Tags  \\\n",
       "11116                                                         <r><dataframe>   \n",
       "44054  <kubernetes><terraform><terraform-provider-aws><iaas><terraform0.12+>   \n",
       "34879                                                     <r><latex><sweave>   \n",
       "43463                                                                    <c>   \n",
       "27023                                                     <javascript><json>   \n",
       "\n",
       "       AnswerCount  CommentCount  FavoriteCount ClosedDate CommunityOwnedDate  \\\n",
       "11116            1             1            NaN        NaN                NaN   \n",
       "44054            1             3            NaN        NaN                NaN   \n",
       "34879            2             0            NaN        NaN                NaN   \n",
       "43463            2             1            NaN        NaN                NaN   \n",
       "27023            2             0            2.0        NaN                NaN   \n",
       "\n",
       "      ContentLicense  \n",
       "11116   CC BY-SA 4.0  \n",
       "44054   CC BY-SA 4.0  \n",
       "34879   CC BY-SA 3.0  \n",
       "43463   CC BY-SA 4.0  \n",
       "27023   CC BY-SA 2.5  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.max_columns\",23)\n",
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "523f842e",
   "metadata": {
    "id": "523f842e"
   },
   "source": [
    "# CARACTERISTIQUES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "03845e71",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 813
    },
    "id": "03845e71",
    "outputId": "d2e13a05-1afa-4fab-e9e5-fe96a4970a77"
   },
   "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>missing_values</th>\n",
       "      <th>%_missing_values</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Id</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PostTypeId</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AcceptedAnswerId</th>\n",
       "      <td>20235</td>\n",
       "      <td>40.470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ParentId</th>\n",
       "      <td>50000</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CreationDate</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DeletionDate</th>\n",
       "      <td>50000</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Score</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ViewCount</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Body</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OwnerUserId</th>\n",
       "      <td>953</td>\n",
       "      <td>1.906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OwnerDisplayName</th>\n",
       "      <td>47214</td>\n",
       "      <td>94.428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LastEditorUserId</th>\n",
       "      <td>22316</td>\n",
       "      <td>44.632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LastEditorDisplayName</th>\n",
       "      <td>48469</td>\n",
       "      <td>96.938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LastEditDate</th>\n",
       "      <td>21867</td>\n",
       "      <td>43.734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LastActivityDate</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Title</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Tags</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AnswerCount</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CommentCount</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FavoriteCount</th>\n",
       "      <td>34828</td>\n",
       "      <td>69.656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ClosedDate</th>\n",
       "      <td>47733</td>\n",
       "      <td>95.466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CommunityOwnedDate</th>\n",
       "      <td>49823</td>\n",
       "      <td>99.646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ContentLicense</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       missing_values  %_missing_values\n",
       "Id                                  0             0.000\n",
       "PostTypeId                          0             0.000\n",
       "AcceptedAnswerId                20235            40.470\n",
       "ParentId                        50000           100.000\n",
       "CreationDate                        0             0.000\n",
       "DeletionDate                    50000           100.000\n",
       "Score                               0             0.000\n",
       "ViewCount                           0             0.000\n",
       "Body                                0             0.000\n",
       "OwnerUserId                       953             1.906\n",
       "OwnerDisplayName                47214            94.428\n",
       "LastEditorUserId                22316            44.632\n",
       "LastEditorDisplayName           48469            96.938\n",
       "LastEditDate                    21867            43.734\n",
       "LastActivityDate                    0             0.000\n",
       "Title                               0             0.000\n",
       "Tags                                0             0.000\n",
       "AnswerCount                         0             0.000\n",
       "CommentCount                        0             0.000\n",
       "FavoriteCount                   34828            69.656\n",
       "ClosedDate                      47733            95.466\n",
       "CommunityOwnedDate              49823            99.646\n",
       "ContentLicense                      0             0.000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfProfil=profil(df)\n",
    "dfProfil[[\"missing_values\",\"%_missing_values\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "4182ff3b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 700
    },
    "id": "4182ff3b",
    "outputId": "a2ac4fe8-8b76-4a0e-e522-14a6a47b5886"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotCategoricalDistributions(plt,df,4,4,10,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "b6c9f831",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 721
    },
    "id": "b6c9f831",
    "outputId": "5539d8ac-f507-4b6b-9f38-3c39a180c09f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotQuantitativeDistributions(plt,df,3,4,10,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "bfb33d92",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "bfb33d92",
    "outputId": "8e606758-7b05-463d-96bf-10433e7fc055",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x1440 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotBoxes(plt,df,3,4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "81daf84e",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 697
    },
    "id": "81daf84e",
    "outputId": "5ca5a392-e5c6-40e2-f347-399f3f0dcb57"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plotCorrelationHeatmap(plt,df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d31efa1",
   "metadata": {
    "id": "6d31efa1"
   },
   "source": [
    "Pour catégoriser les questions, nous utiliserons les colonnes title, body et tag car ces colonnes contiennent du texte et n'ont pas de valeurs manquantes.  \n",
    "Les colonnes OwnerDisplayName et LastEditorDisplayName auraient pu être intéressantes mais elles ont toutes les deux plus de 94% de valeurs manquantes.  \n",
    "Nous ne les garderons donc pas."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ef996e8",
   "metadata": {
    "id": "2ef996e8"
   },
   "source": [
    "# NETTOYAGE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "974d7653",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 752
    },
    "id": "974d7653",
    "outputId": "1ecb0d0e-e492-4ba9-a4e0-8689d975cf84"
   },
   "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>Title</th>\n",
       "      <th>Body</th>\n",
       "      <th>Tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>comprehensive image processing example using cocoa api</td>\n",
       "      <td>&lt;p&gt;can someone point out a comprehensive example on image processing using cocoa api? i am developing an application...</td>\n",
       "      <td>&lt;cocoa&gt;&lt;image-processing&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>testing gps in iphone simulator problem</td>\n",
       "      <td>&lt;p&gt;i found some various articles about testing the gps (corelocation) in the iphone simulator. \\nit seems pretty str...</td>\n",
       "      <td>&lt;iphone&gt;&lt;gps&gt;&lt;core-location&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>create pdf inapp ios</td>\n",
       "      <td>&lt;blockquote&gt;\\n  &lt;p&gt;&lt;strong&gt;possible duplicate:&lt;/strong&gt;&lt;br&gt;\\n  &lt;a href=\"https://stackoverflow.com/questions/1360912/...</td>\n",
       "      <td>&lt;ios&gt;&lt;pdf&gt;&lt;pdf-generation&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>flex 4 stopping an embbeded swf</td>\n",
       "      <td>&lt;p&gt;i have this flex 4.1 script:&lt;/p&gt;\\n\\n&lt;pre&gt;&lt;code&gt;&amp;lt;fx:script&amp;gt;\\n    &amp;lt;![cdata[\\n        [embed(source=\"res/sw...</td>\n",
       "      <td>&lt;apache-flex&gt;&lt;actionscript-3&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>html form - firing the form submit using the enter key does not work when focus is on a select list</td>\n",
       "      <td>&lt;p&gt;i'm stuck on the above problem. i have a simple form as follows, with a text input, a select list and a submit bu...</td>\n",
       "      <td>&lt;html&gt;&lt;forms&gt;&lt;submit&gt;</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                 Title  \\\n",
       "0                                               comprehensive image processing example using cocoa api   \n",
       "1                                                              testing gps in iphone simulator problem   \n",
       "2                                                                                 create pdf inapp ios   \n",
       "3                                                                      flex 4 stopping an embbeded swf   \n",
       "4  html form - firing the form submit using the enter key does not work when focus is on a select list   \n",
       "\n",
       "                                                                                                                      Body  \\\n",
       "0  <p>can someone point out a comprehensive example on image processing using cocoa api? i am developing an application...   \n",
       "1  <p>i found some various articles about testing the gps (corelocation) in the iphone simulator. \\nit seems pretty str...   \n",
       "2  <blockquote>\\n  <p><strong>possible duplicate:</strong><br>\\n  <a href=\"https://stackoverflow.com/questions/1360912/...   \n",
       "3  <p>i have this flex 4.1 script:</p>\\n\\n<pre><code>&lt;fx:script&gt;\\n    &lt;![cdata[\\n        [embed(source=\"res/sw...   \n",
       "4  <p>i'm stuck on the above problem. i have a simple form as follows, with a text input, a select list and a submit bu...   \n",
       "\n",
       "                            Tags  \n",
       "0      <cocoa><image-processing>  \n",
       "1   <iphone><gps><core-location>  \n",
       "2     <ios><pdf><pdf-generation>  \n",
       "3  <apache-flex><actionscript-3>  \n",
       "4          <html><forms><submit>  "
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# on ne garde que les colonnes contenant du texte\n",
    "dfCleaned=df[['Title','Body','Tags']]\n",
    "\n",
    "# uniformisation du dataframe en minuscule\n",
    "dfCleaned=dfCleaned.astype(str).apply(lambda x: x.str.lower())\n",
    "# aperçu des données\n",
    "dfCleaned.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "5bc05ce6",
   "metadata": {
    "id": "5bc05ce6"
   },
   "outputs": [],
   "source": [
    "# Titre: suppression des caracteres speciaux et ponctuations\n",
    "dfCleaned['CleanedTitle']=dfCleaned['Title'].replace('[\"!\\\"#$%&\\(\\)\\*\\+-\\.\\/:;<=>\\?\\@\\\\^_`\\{\\|\\}~\\n\"]',' ',regex=True).astype(str)\n",
    "\n",
    "# Corps: suppression des balises html, \\n,;,//,date ou chiffres,chaine contenant le caractere _, ponctuations  dans le Body\n",
    "dfCleaned['CleanedBody']=dfCleaned['Body'].replace('(<\\/?[a-z]+>|&lt;|&gt;|//|[\"!\\\"#$%&\\(\\)\\*\\+-\\.\\/:;<=>\\?\\@\\\\^_\\'`\\{\\|\\}~\\n\"]|\\d+.\\d+|\\S+_\\S+|\\*+|[0-9]+[a-z]+|[\\d+]|[abcd]{2,}[a-z]*)', ' ',regex=True).astype(str)\n",
    "#suppression des blancs superflus\n",
    "dfCleaned['CleanedBody']=dfCleaned['CleanedBody'].str.strip()\n",
    "\n",
    "# suppression des <,>  dans les Tags\n",
    "dfCleaned['Tags']=dfCleaned['Tags'].replace('[<>]', ' ',regex=True).astype(str)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e834cc0",
   "metadata": {
    "id": "1e834cc0"
   },
   "source": [
    "## Nettoyage des stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "8514079b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8514079b",
    "outputId": "645f56a1-b11d-4f9a-8779-56ee1d4ae9a8",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3.42 s, sys: 16.9 ms, total: 3.44 s\n",
      "Wall time: 3.44 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "#  suppression des stopwords dans le body, alimente la colonne FilteredBody\n",
    "stop_words = set(stopwords.words('english'))\n",
    "\n",
    "filteredBody=[]\n",
    "for i in range(len(dfCleaned['CleanedBody'])):\n",
    "    filteredBody.append(' '.join([word for word in re.split(r\"\\W+\", dfCleaned['CleanedBody'][i]) if word not in stop_words]))\n",
    "dfCleaned['FilteredBody']=filteredBody\n",
    "\n",
    "filteredTitle=[]\n",
    "for i in range(len(dfCleaned['CleanedTitle'])):\n",
    "    filteredTitle.append(' '.join([word for word in re.split(r\"\\W+\", dfCleaned['CleanedTitle'][i]) if word not in stop_words]))\n",
    "dfCleaned['FilteredTitle']=filteredTitle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "d9ceef88",
   "metadata": {
    "id": "d9ceef88"
   },
   "outputs": [],
   "source": [
    "del dfCleaned['CleanedBody']\n",
    "del dfCleaned['CleanedTitle']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "8dbc18a2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 580
    },
    "id": "8dbc18a2",
    "outputId": "6cc076ea-c4f1-4f1b-a42a-03ac99f0383c",
    "scrolled": 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>Title</th>\n",
       "      <th>Body</th>\n",
       "      <th>Tags</th>\n",
       "      <th>FilteredBody</th>\n",
       "      <th>FilteredTitle</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>526</th>\n",
       "      <td>entity class from database....error invalid resource _pm</td>\n",
       "      <td>&lt;p&gt;i have created a database and table in mysql through phpmyadmin. i am trying to connect this to my project using ...</td>\n",
       "      <td>mysql  jdbc  netbeans  entity</td>\n",
       "      <td>created mysql phpmy trying connect project using entity class however get following error run project com sun appser...</td>\n",
       "      <td>entity class database error invalid resource pm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48647</th>\n",
       "      <td>formatting using decimalformat throws exception - \"cannot format given object as a number\"</td>\n",
       "      <td>&lt;p&gt;this might look like a repeated question but i tried in all the below links and can't get a proper answer.&lt;/p&gt;\\n\\...</td>\n",
       "      <td>java  exception  overloading  illegalargumentexception  decimalformat</td>\n",
       "      <td>might look like repeated question tried links get proper answer href https st com questions format given object numb...</td>\n",
       "      <td>formatting using decimalformat throws exception cannot format given object number</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32283</th>\n",
       "      <td>jsf links not working on facelets</td>\n",
       "      <td>&lt;p&gt;i have created a jsf 2 page which uses facelets to define the structure of the page.&lt;/p&gt;\\n\\n&lt;pre&gt;&lt;code&gt;&amp;lt;div cl...</td>\n",
       "      <td>jsf  jakarta-ee  jsf-2  facelets</td>\n",
       "      <td>created jsf page uses f define structure page div class page div class content ui insert name content ui insert div ...</td>\n",
       "      <td>jsf links working facelets</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8080</th>\n",
       "      <td>angularjs - ng-disabled doesn't work as expected</td>\n",
       "      <td>&lt;p&gt;i am trying to disable file &lt;code&gt;input&lt;/code&gt; tag, after user selected a file.&lt;/p&gt;\\n\\n&lt;p&gt;html:&lt;/p&gt;\\n\\n&lt;pre&gt;&lt;code...</td>\n",
       "      <td>javascript  html  angularjs</td>\n",
       "      <td>trying dis file input tag user selected file html div ng controller firstcontroller div ng controller secon input ty...</td>\n",
       "      <td>angularjs ng disabled work expected</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12463</th>\n",
       "      <td>split string delimited by comma without respect to commas in brackets</td>\n",
       "      <td>&lt;p&gt;i've got a string like&lt;/p&gt;\\n\\n&lt;pre&gt;&lt;code&gt;s=\"abc, 3rncd (23uh, sdfuh), 32h(q23q)89 (as), dwe8h, edt (1,wer,345,rtz...</td>\n",
       "      <td>java  regex  string  split</td>\n",
       "      <td>got string like sdfuh q dwe edt wer rtz tr nope want split string string parts sdfuh q dwe edt wer rtz tr nope simpl...</td>\n",
       "      <td>split string delimited comma without respect commas brackets</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43886</th>\n",
       "      <td>nant merge task not executing</td>\n",
       "      <td>&lt;p&gt;i have a nant merge task aimed at a nunit test output.xml  i have installed the relevant xsl transform files and ...</td>\n",
       "      <td>.net  cruisecontrol.net  nant</td>\n",
       "      <td>nant merge task aimed nunit test output xml installed relevant xsl transform files menu option show see nunit detail...</td>\n",
       "      <td>nant merge task executing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40164</th>\n",
       "      <td>delphi rtti get property - why does this results in av?</td>\n",
       "      <td>&lt;p&gt;i am trying to write a spec utility library.&lt;/p&gt;\\n\\n&lt;p&gt;one of the specification is a texpressionspecification. ba...</td>\n",
       "      <td>delphi  generics  rtti</td>\n",
       "      <td>trying write spec utility library one specifi texpressionspecifi implements specifi pattern evaluating inner texpres...</td>\n",
       "      <td>delphi rtti get property results av</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10440</th>\n",
       "      <td>arduino - about data types</td>\n",
       "      <td>&lt;p&gt;i am writing a delay function for arduino to be used in a project. the reason is that timer0 (the default timer u...</td>\n",
       "      <td>c++  arduino-uno</td>\n",
       "      <td>writing delay function arduino used project reason timer default timer used arduino delay pwm control test code dela...</td>\n",
       "      <td>arduino data types</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10531</th>\n",
       "      <td>is there any blog plugin which i can use with my existing rails3 application?</td>\n",
       "      <td>&lt;p&gt;i need to add a blog to my existing rails application. i am looking for a such blog-plugin which also give me abi...</td>\n",
       "      <td>ruby-on-rails  ruby-on-rails-3  plugins</td>\n",
       "      <td>need blog existing rails appli looking blog plugin also give tagging blogs seo friendly urls kind plugin gem avail c...</td>\n",
       "      <td>blog plugin use existing rails3 application</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4052</th>\n",
       "      <td>update some chrome extension files from own server</td>\n",
       "      <td>&lt;p&gt;i am writing a chrome extension in which some script and an xml file in which i data is to be read by the script ...</td>\n",
       "      <td>google-chrome-extension  auto-update</td>\n",
       "      <td>writing chrome extension script xml file script changed almost twice per include link every page lo hosting server l...</td>\n",
       "      <td>update chrome extension files server</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                            Title  \\\n",
       "526                                      entity class from database....error invalid resource _pm   \n",
       "48647  formatting using decimalformat throws exception - \"cannot format given object as a number\"   \n",
       "32283                                                           jsf links not working on facelets   \n",
       "8080                                             angularjs - ng-disabled doesn't work as expected   \n",
       "12463                       split string delimited by comma without respect to commas in brackets   \n",
       "43886                                                               nant merge task not executing   \n",
       "40164                                     delphi rtti get property - why does this results in av?   \n",
       "10440                                                                  arduino - about data types   \n",
       "10531               is there any blog plugin which i can use with my existing rails3 application?   \n",
       "4052                                           update some chrome extension files from own server   \n",
       "\n",
       "                                                                                                                          Body  \\\n",
       "526    <p>i have created a database and table in mysql through phpmyadmin. i am trying to connect this to my project using ...   \n",
       "48647  <p>this might look like a repeated question but i tried in all the below links and can't get a proper answer.</p>\\n\\...   \n",
       "32283  <p>i have created a jsf 2 page which uses facelets to define the structure of the page.</p>\\n\\n<pre><code>&lt;div cl...   \n",
       "8080   <p>i am trying to disable file <code>input</code> tag, after user selected a file.</p>\\n\\n<p>html:</p>\\n\\n<pre><code...   \n",
       "12463  <p>i've got a string like</p>\\n\\n<pre><code>s=\"abc, 3rncd (23uh, sdfuh), 32h(q23q)89 (as), dwe8h, edt (1,wer,345,rtz...   \n",
       "43886  <p>i have a nant merge task aimed at a nunit test output.xml  i have installed the relevant xsl transform files and ...   \n",
       "40164  <p>i am trying to write a spec utility library.</p>\\n\\n<p>one of the specification is a texpressionspecification. ba...   \n",
       "10440  <p>i am writing a delay function for arduino to be used in a project. the reason is that timer0 (the default timer u...   \n",
       "10531  <p>i need to add a blog to my existing rails application. i am looking for a such blog-plugin which also give me abi...   \n",
       "4052   <p>i am writing a chrome extension in which some script and an xml file in which i data is to be read by the script ...   \n",
       "\n",
       "                                                                          Tags  \\\n",
       "526                                             mysql  jdbc  netbeans  entity    \n",
       "48647   java  exception  overloading  illegalargumentexception  decimalformat    \n",
       "32283                                        jsf  jakarta-ee  jsf-2  facelets    \n",
       "8080                                              javascript  html  angularjs    \n",
       "12463                                              java  regex  string  split    \n",
       "43886                                           .net  cruisecontrol.net  nant    \n",
       "40164                                                  delphi  generics  rtti    \n",
       "10440                                                        c++  arduino-uno    \n",
       "10531                                 ruby-on-rails  ruby-on-rails-3  plugins    \n",
       "4052                                     google-chrome-extension  auto-update    \n",
       "\n",
       "                                                                                                                  FilteredBody  \\\n",
       "526    created mysql phpmy trying connect project using entity class however get following error run project com sun appser...   \n",
       "48647  might look like repeated question tried links get proper answer href https st com questions format given object numb...   \n",
       "32283  created jsf page uses f define structure page div class page div class content ui insert name content ui insert div ...   \n",
       "8080   trying dis file input tag user selected file html div ng controller firstcontroller div ng controller secon input ty...   \n",
       "12463  got string like sdfuh q dwe edt wer rtz tr nope want split string string parts sdfuh q dwe edt wer rtz tr nope simpl...   \n",
       "43886  nant merge task aimed nunit test output xml installed relevant xsl transform files menu option show see nunit detail...   \n",
       "40164  trying write spec utility library one specifi texpressionspecifi implements specifi pattern evaluating inner texpres...   \n",
       "10440  writing delay function arduino used project reason timer default timer used arduino delay pwm control test code dela...   \n",
       "10531  need blog existing rails appli looking blog plugin also give tagging blogs seo friendly urls kind plugin gem avail c...   \n",
       "4052   writing chrome extension script xml file script changed almost twice per include link every page lo hosting server l...   \n",
       "\n",
       "                                                                            FilteredTitle  \n",
       "526                                       entity class database error invalid resource pm  \n",
       "48647  formatting using decimalformat throws exception cannot format given object number   \n",
       "32283                                                          jsf links working facelets  \n",
       "8080                                                  angularjs ng disabled work expected  \n",
       "12463                        split string delimited comma without respect commas brackets  \n",
       "43886                                                           nant merge task executing  \n",
       "40164                                                delphi rtti get property results av   \n",
       "10440                                                                  arduino data types  \n",
       "10531                                        blog plugin use existing rails3 application   \n",
       "4052                                                 update chrome extension files server  "
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfCleaned.sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb7fcf90",
   "metadata": {
    "id": "10425915"
   },
   "source": [
    "## Tokenisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "3e97c339",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.35 s, sys: 130 ms, total: 1.48 s\n",
      "Wall time: 1.48 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "# conversion du corpus de chaque document en tokens unique\n",
    "dfCleaned['BodyTokens']=dfCleaned['FilteredBody'].map(tokenize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "04cb0823",
   "metadata": {},
   "outputs": [],
   "source": [
    "dfCleaned['TitleTokens']=dfCleaned['FilteredTitle'].map(tokenize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "7813a4e6",
   "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>Title</th>\n",
       "      <th>Body</th>\n",
       "      <th>Tags</th>\n",
       "      <th>FilteredBody</th>\n",
       "      <th>FilteredTitle</th>\n",
       "      <th>BodyTokens</th>\n",
       "      <th>TitleTokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30057</th>\n",
       "      <td>swap arrays by using pointers in c++</td>\n",
       "      <td>&lt;p&gt;i have two arrays of pointers to doubles that i need to swap. rather than just copy the data within the arrays, i...</td>\n",
       "      <td>c++  arrays  pointers  swap</td>\n",
       "      <td>two arrays pointers doubles need swap rather copy within arrays would efficient swap pointers arrays always impressi...</td>\n",
       "      <td>swap arrays using pointers c</td>\n",
       "      <td>[two, arrays, pointers, doubles, need, swap, rather, copy, within, arrays, would, efficient, swap, pointers, arrays,...</td>\n",
       "      <td>[swap, arrays, using, pointers]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17182</th>\n",
       "      <td>heroku 409 error when python dash app calls api</td>\n",
       "      <td>&lt;p&gt;starting process with command from procfile works.&lt;/p&gt;\\n\\n&lt;p&gt;then as soon as it starts reading the python file, a...</td>\n",
       "      <td>python  api  heroku  server  plotly-dash</td>\n",
       "      <td>starting process command procfile works soon starts python file pulling api get error exception worker process tr ra...</td>\n",
       "      <td>heroku 409 error python dash app calls api</td>\n",
       "      <td>[starting, process, command, procfile, works, soon, starts, python, file, pulling, api, get, error, exception, worke...</td>\n",
       "      <td>[heroku, 409, error, python, dash, app, calls, api]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5076</th>\n",
       "      <td>set a usercontrol property to not show up in vs properties window</td>\n",
       "      <td>&lt;p&gt;i have a usercontrol in my asp.net project that has a public property. i do not want this property to show up in ...</td>\n",
       "      <td>c#  asp.net  visual-studio  properties  attributes</td>\n",
       "      <td>usercontrol asp net project public property want property show visual studio property window user highlights instanc...</td>\n",
       "      <td>set usercontrol property show vs properties window</td>\n",
       "      <td>[usercontrol, asp, net, project, public, property, want, property, show, visual, studio, property, window, user, hig...</td>\n",
       "      <td>[set, usercontrol, property, show, vs, properties, window]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12878</th>\n",
       "      <td>how can we define one-to-one, one-to-many, many-to-one, and many-to-many lstm neural networks in keras?</td>\n",
       "      <td>&lt;p&gt;i am reading &lt;a href=\"http://karpathy.github.io/2015/05/21/rnn-effectiveness/\" rel=\"noreferrer\"&gt;this&lt;/a&gt; article ...</td>\n",
       "      <td>python  neural-network  keras  lstm  recurrent-neural-network</td>\n",
       "      <td>href http karpathy github io rnn effectiveness rel noreferrer article unreason effectiveness recurrent neural networ...</td>\n",
       "      <td>define one one one many many one many many lstm neural networks keras</td>\n",
       "      <td>[href, http, karpathy, github, io, rnn, effectiveness, rel, noreferrer, article, unreason, effectiveness, recurrent,...</td>\n",
       "      <td>[define, one, one, one, many, many, one, many, many, lstm, neural, networks, keras]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27534</th>\n",
       "      <td>is it necessary to upgrade to jquery 1.4?</td>\n",
       "      <td>&lt;p&gt;i have been using jquery 1.23 till now and nowadays most of the plugins are being developed for the 1.4 version. ...</td>\n",
       "      <td>jquery</td>\n",
       "      <td>using jquery till plugins developed version want know necessary upgr jquery mean modules right still work</td>\n",
       "      <td>necessary upgrade jquery 1 4</td>\n",
       "      <td>[using, jquery, till, plugins, developed, version, want, know, necessary, upgr, jquery, mean, modules, right, still,...</td>\n",
       "      <td>[necessary, upgrade, jquery]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21290</th>\n",
       "      <td>making a class behave like a bunch of keyword arguments in python</td>\n",
       "      <td>&lt;p&gt;so i have a class that will have a lot of attributes. is there a way to make this class behave in a way that i ca...</td>\n",
       "      <td>python</td>\n",
       "      <td>class lot attributes way make class behave way pass kwarg function class foo object def self self self b b foo foo s...</td>\n",
       "      <td>making class behave like bunch keyword arguments python</td>\n",
       "      <td>[class, lot, attributes, way, make, class, behave, way, pass, kwarg, function, class, foo, object, def, self, self, ...</td>\n",
       "      <td>[making, class, behave, like, bunch, keyword, arguments, python]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33326</th>\n",
       "      <td>output gzipped result as a stream</td>\n",
       "      <td>&lt;p&gt;i have a simple php script that output a sitemap from a list of urls. &lt;/p&gt;\\n\\n&lt;p&gt;the sitemap is generated on the ...</td>\n",
       "      <td>php  gzip</td>\n",
       "      <td>simple php script output sitemap list urls sitemap generated fly far output xml directly works think compressed xml ...</td>\n",
       "      <td>output gzipped result stream</td>\n",
       "      <td>[simple, php, script, output, sitemap, list, urls, sitemap, generated, fly, far, output, xml, directly, works, think...</td>\n",
       "      <td>[output, gzipped, result, stream]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45971</th>\n",
       "      <td>java.lang.classcastexception: org.jboss.ws.core.soap.textimpl cannot be cast to javax.xml.soap.soapelement</td>\n",
       "      <td>&lt;p&gt;i am working on &lt;strong&gt;ibm sterling v 9.3&lt;/strong&gt;. i have exposed an service as &lt;strong&gt;jax-ws web service&lt;/str...</td>\n",
       "      <td>web-services  jboss  jax-ws  soapui  ibm-sterling</td>\n",
       "      <td>working ibm sterling v exposed service jax ws web service using steps provided knowledge center su jboss app server ...</td>\n",
       "      <td>java lang classcastexception org jboss ws core soap textimpl cannot cast javax xml soap soapelement</td>\n",
       "      <td>[working, ibm, sterling, exposed, service, jax, ws, web, service, using, steps, provided, knowledge, center, su, jbo...</td>\n",
       "      <td>[java, lang, classcastexception, org, jboss, ws, core, soap, textimpl, cannot, cast, javax, xml, soap, soapelement]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                            Title  \\\n",
       "30057                                                                        swap arrays by using pointers in c++   \n",
       "17182                                                             heroku 409 error when python dash app calls api   \n",
       "5076                                            set a usercontrol property to not show up in vs properties window   \n",
       "12878     how can we define one-to-one, one-to-many, many-to-one, and many-to-many lstm neural networks in keras?   \n",
       "27534                                                                   is it necessary to upgrade to jquery 1.4?   \n",
       "21290                                           making a class behave like a bunch of keyword arguments in python   \n",
       "33326                                                                           output gzipped result as a stream   \n",
       "45971  java.lang.classcastexception: org.jboss.ws.core.soap.textimpl cannot be cast to javax.xml.soap.soapelement   \n",
       "\n",
       "                                                                                                                          Body  \\\n",
       "30057  <p>i have two arrays of pointers to doubles that i need to swap. rather than just copy the data within the arrays, i...   \n",
       "17182  <p>starting process with command from procfile works.</p>\\n\\n<p>then as soon as it starts reading the python file, a...   \n",
       "5076   <p>i have a usercontrol in my asp.net project that has a public property. i do not want this property to show up in ...   \n",
       "12878  <p>i am reading <a href=\"http://karpathy.github.io/2015/05/21/rnn-effectiveness/\" rel=\"noreferrer\">this</a> article ...   \n",
       "27534  <p>i have been using jquery 1.23 till now and nowadays most of the plugins are being developed for the 1.4 version. ...   \n",
       "21290  <p>so i have a class that will have a lot of attributes. is there a way to make this class behave in a way that i ca...   \n",
       "33326  <p>i have a simple php script that output a sitemap from a list of urls. </p>\\n\\n<p>the sitemap is generated on the ...   \n",
       "45971  <p>i am working on <strong>ibm sterling v 9.3</strong>. i have exposed an service as <strong>jax-ws web service</str...   \n",
       "\n",
       "                                                                  Tags  \\\n",
       "30057                                     c++  arrays  pointers  swap    \n",
       "17182                        python  api  heroku  server  plotly-dash    \n",
       "5076               c#  asp.net  visual-studio  properties  attributes    \n",
       "12878   python  neural-network  keras  lstm  recurrent-neural-network    \n",
       "27534                                                          jquery    \n",
       "21290                                                          python    \n",
       "33326                                                       php  gzip    \n",
       "45971               web-services  jboss  jax-ws  soapui  ibm-sterling    \n",
       "\n",
       "                                                                                                                  FilteredBody  \\\n",
       "30057  two arrays pointers doubles need swap rather copy within arrays would efficient swap pointers arrays always impressi...   \n",
       "17182  starting process command procfile works soon starts python file pulling api get error exception worker process tr ra...   \n",
       "5076   usercontrol asp net project public property want property show visual studio property window user highlights instanc...   \n",
       "12878  href http karpathy github io rnn effectiveness rel noreferrer article unreason effectiveness recurrent neural networ...   \n",
       "27534                using jquery till plugins developed version want know necessary upgr jquery mean modules right still work   \n",
       "21290  class lot attributes way make class behave way pass kwarg function class foo object def self self self b b foo foo s...   \n",
       "33326  simple php script output sitemap list urls sitemap generated fly far output xml directly works think compressed xml ...   \n",
       "45971  working ibm sterling v exposed service jax ws web service using steps provided knowledge center su jboss app server ...   \n",
       "\n",
       "                                                                                             FilteredTitle  \\\n",
       "30057                                                                        swap arrays using pointers c    \n",
       "17182                                                           heroku 409 error python dash app calls api   \n",
       "5076                                                    set usercontrol property show vs properties window   \n",
       "12878                               define one one one many many one many many lstm neural networks keras    \n",
       "27534                                                                        necessary upgrade jquery 1 4    \n",
       "21290                                              making class behave like bunch keyword arguments python   \n",
       "33326                                                                         output gzipped result stream   \n",
       "45971  java lang classcastexception org jboss ws core soap textimpl cannot cast javax xml soap soapelement   \n",
       "\n",
       "                                                                                                                    BodyTokens  \\\n",
       "30057  [two, arrays, pointers, doubles, need, swap, rather, copy, within, arrays, would, efficient, swap, pointers, arrays,...   \n",
       "17182  [starting, process, command, procfile, works, soon, starts, python, file, pulling, api, get, error, exception, worke...   \n",
       "5076   [usercontrol, asp, net, project, public, property, want, property, show, visual, studio, property, window, user, hig...   \n",
       "12878  [href, http, karpathy, github, io, rnn, effectiveness, rel, noreferrer, article, unreason, effectiveness, recurrent,...   \n",
       "27534  [using, jquery, till, plugins, developed, version, want, know, necessary, upgr, jquery, mean, modules, right, still,...   \n",
       "21290  [class, lot, attributes, way, make, class, behave, way, pass, kwarg, function, class, foo, object, def, self, self, ...   \n",
       "33326  [simple, php, script, output, sitemap, list, urls, sitemap, generated, fly, far, output, xml, directly, works, think...   \n",
       "45971  [working, ibm, sterling, exposed, service, jax, ws, web, service, using, steps, provided, knowledge, center, su, jbo...   \n",
       "\n",
       "                                                                                                               TitleTokens  \n",
       "30057                                                                                      [swap, arrays, using, pointers]  \n",
       "17182                                                                  [heroku, 409, error, python, dash, app, calls, api]  \n",
       "5076                                                            [set, usercontrol, property, show, vs, properties, window]  \n",
       "12878                                  [define, one, one, one, many, many, one, many, many, lstm, neural, networks, keras]  \n",
       "27534                                                                                         [necessary, upgrade, jquery]  \n",
       "21290                                                     [making, class, behave, like, bunch, keyword, arguments, python]  \n",
       "33326                                                                                    [output, gzipped, result, stream]  \n",
       "45971  [java, lang, classcastexception, org, jboss, ws, core, soap, textimpl, cannot, cast, javax, xml, soap, soapelement]  "
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfCleaned.sample(8)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "811c7269",
   "metadata": {
    "id": "811c7269"
   },
   "source": [
    "## Lemmatisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "9fce225f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 8min 20s, sys: 1min 14s, total: 9min 35s\n",
      "Wall time: 9min 36s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "dfCleaned['BodyLemmas']=dfCleaned['BodyTokens'].map(lemmatize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "93d024d5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 31.4 s, sys: 4.62 s, total: 36.1 s\n",
      "Wall time: 36.1 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "dfCleaned['TitleLemmas']=dfCleaned['TitleTokens'].map(lemmatize)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "6dbf585d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def highlight_cols(s):\n",
    "    color = 'lightgrey'\n",
    "    return 'background-color: %s' % color"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "be89c3e9",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_25042_row0_col0, #T_25042_row0_col1, #T_25042_row0_col2, #T_25042_row1_col0, #T_25042_row1_col1, #T_25042_row1_col2, #T_25042_row2_col0, #T_25042_row2_col1, #T_25042_row2_col2 {\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_25042_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >Title</th>\n",
       "      <th class=\"col_heading level0 col1\" >Body</th>\n",
       "      <th class=\"col_heading level0 col2\" >Tags</th>\n",
       "      <th class=\"col_heading level0 col3\" >FilteredBody</th>\n",
       "      <th class=\"col_heading level0 col4\" >FilteredTitle</th>\n",
       "      <th class=\"col_heading level0 col5\" >BodyTokens</th>\n",
       "      <th class=\"col_heading level0 col6\" >TitleTokens</th>\n",
       "      <th class=\"col_heading level0 col7\" >BodyLemmas</th>\n",
       "      <th class=\"col_heading level0 col8\" >TitleLemmas</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_25042_level0_row0\" class=\"row_heading level0 row0\" >45583</th>\n",
       "      <td id=\"T_25042_row0_col0\" class=\"data row0 col0\" >wordpress > create category as child of \"uncategorized\" (catid=1)</td>\n",
       "      <td id=\"T_25042_row0_col1\" class=\"data row0 col1\" ><p>i've got a script in my functions.php file that checks for the existence of several categories that are required by my theme, and if they are not present, it creates them...</p>\n",
       "\n",
       "<pre><code>if(!get_cat_id('my-category')){wp_create_category('my-category');}\n",
       "</code></pre>\n",
       "\n",
       "<p>i need to modify this script in two ways...</p>\n",
       "\n",
       "<p>first, i need to create the category so that its parent category is the \"uncategorized\" category (or id 1).</p>\n",
       "\n",
       "<p>second, if the category already exists, but its parent is not id 1, i need to make it so.</p>\n",
       "</td>\n",
       "      <td id=\"T_25042_row0_col2\" class=\"data row0 col2\" > wordpress  categories </td>\n",
       "      <td id=\"T_25042_row0_col3\" class=\"data row0 col3\" >got script functions php file checks existence several required theme present creates need modify script two ways first need create parent un id second alre exists parent id need make</td>\n",
       "      <td id=\"T_25042_row0_col4\" class=\"data row0 col4\" >wordpress create category child uncategorized catid 1 </td>\n",
       "      <td id=\"T_25042_row0_col5\" class=\"data row0 col5\" >['got', 'script', 'functions', 'php', 'file', 'checks', 'existence', 'several', 'required', 'theme', 'present', 'creates', 'need', 'modify', 'script', 'two', 'ways', 'first', 'need', 'create', 'parent', 'un', 'id', 'second', 'alre', 'exists', 'parent', 'id', 'need', 'make']</td>\n",
       "      <td id=\"T_25042_row0_col6\" class=\"data row0 col6\" >['wordpress', 'create', 'category', 'child', 'uncategorized', 'catid']</td>\n",
       "      <td id=\"T_25042_row0_col7\" class=\"data row0 col7\" >['get', 'script', 'function', 'php', 'file', 'check', 'existence', 'several', 'require', 'theme', 'present', 'creates', 'need', 'modify', 'script', 'two', 'way', 'first', 'need', 'create', 'parent', 'un', 'id', 'second', 'alre', 'exists', 'parent', 'id', 'need', 'make']</td>\n",
       "      <td id=\"T_25042_row0_col8\" class=\"data row0 col8\" >['wordpress', 'create', 'category', 'child', 'uncategorized', 'catid']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_25042_level0_row1\" class=\"row_heading level0 row1\" >23834</th>\n",
       "      <td id=\"T_25042_row1_col0\" class=\"data row1 col0\" >how to reshape matlab matrices for this example?</td>\n",
       "      <td id=\"T_25042_row1_col1\" class=\"data row1 col1\" ><p>i have a 40x16 matrix or 8 5x16 one below the other i.e. aligned vertically. \n",
       "i want to get a 5x128 matrix from that such that i align the 8 5x16 matrices horizontally.\n",
       "is there an efficient/quicker (rather than the hardcoded for loops) way to do this?</p>\n",
       "\n",
       "<p>i want the individual 5x16 matrices intact.</p>\n",
       "</td>\n",
       "      <td id=\"T_25042_row1_col2\" class=\"data row1 col2\" > matlab  matrix  reshape </td>\n",
       "      <td id=\"T_25042_row1_col3\" class=\"data row1 col3\" >matrix x one e aligned verti want get matrix align x matrices horizontally efficient quicker rather har loops way want individual matrices int</td>\n",
       "      <td id=\"T_25042_row1_col4\" class=\"data row1 col4\" >reshape matlab matrices example </td>\n",
       "      <td id=\"T_25042_row1_col5\" class=\"data row1 col5\" >['matrix', 'one', 'aligned', 'verti', 'want', 'get', 'matrix', 'align', 'matrices', 'horizontally', 'efficient', 'quicker', 'rather', 'har', 'loops', 'way', 'want', 'individual', 'matrices', 'int']</td>\n",
       "      <td id=\"T_25042_row1_col6\" class=\"data row1 col6\" >['reshape', 'matlab', 'matrices', 'example']</td>\n",
       "      <td id=\"T_25042_row1_col7\" class=\"data row1 col7\" >['matrix', 'one', 'align', 'verti', 'want', 'get', 'matrix', 'align', 'matrix', 'horizontally', 'efficient', 'quicker', 'rather', 'har', 'loop', 'way', 'want', 'individual', 'matrix', 'int']</td>\n",
       "      <td id=\"T_25042_row1_col8\" class=\"data row1 col8\" >['reshape', 'matlab', 'matrix', 'example']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_25042_level0_row2\" class=\"row_heading level0 row2\" >24110</th>\n",
       "      <td id=\"T_25042_row2_col0\" class=\"data row2 col0\" >regex syntax validator</td>\n",
       "      <td id=\"T_25042_row2_col1\" class=\"data row2 col1\" ><p>does anybody know any tool for validating the syntax of an regular expression? i dont want to validate if it matches or not with some text but i want to see if there are syntax errors in the regex (missing parenthisis etc).</p>\n",
       "\n",
       "<p>also, what about syntax highlighting? it would be a great help when writing complex regex.</p>\n",
       "</td>\n",
       "      <td id=\"T_25042_row2_col2\" class=\"data row2 col2\" > regex  syntax  highlighting  validation </td>\n",
       "      <td id=\"T_25042_row2_col3\" class=\"data row2 col3\" >anybody know tool vali syntax regular expression dont want vali matches text want see syntax errors regex missing parenthisis etc also syntax highlighting would great help writing complex regex</td>\n",
       "      <td id=\"T_25042_row2_col4\" class=\"data row2 col4\" >regex syntax validator</td>\n",
       "      <td id=\"T_25042_row2_col5\" class=\"data row2 col5\" >['anybody', 'know', 'tool', 'vali', 'syntax', 'regular', 'expression', 'dont', 'want', 'vali', 'matches', 'text', 'want', 'see', 'syntax', 'errors', 'regex', 'missing', 'parenthisis', 'etc', 'also', 'syntax', 'highlighting', 'would', 'great', 'help', 'writing', 'complex', 'regex']</td>\n",
       "      <td id=\"T_25042_row2_col6\" class=\"data row2 col6\" >['regex', 'syntax', 'validator']</td>\n",
       "      <td id=\"T_25042_row2_col7\" class=\"data row2 col7\" >['anybody', 'know', 'tool', 'vali', 'syntax', 'regular', 'expression', 'dont', 'want', 'vali', 'match', 'text', 'want', 'see', 'syntax', 'error', 'regex', 'miss', 'parenthisis', 'etc', 'also', 'syntax', 'highlight', 'would', 'great', 'help', 'write', 'complex', 'regex']</td>\n",
       "      <td id=\"T_25042_row2_col8\" class=\"data row2 col8\" >['regex', 'syntax', 'validator']</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fe15a76e4f0>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfCleaned.sample(3).style.applymap(highlight_cols, subset=pd.IndexSlice[:, ['Title', 'Body','Tags']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "110f4f20",
   "metadata": {
    "scrolled": 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>Title</th>\n",
       "      <th>Body</th>\n",
       "      <th>Tags</th>\n",
       "      <th>FilteredBody</th>\n",
       "      <th>FilteredTitle</th>\n",
       "      <th>BodyTokens</th>\n",
       "      <th>TitleTokens</th>\n",
       "      <th>BodyLemmas</th>\n",
       "      <th>TitleLemmas</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14013</th>\n",
       "      <td>add custom field inside a do_shortcode</td>\n",
       "      <td>&lt;p&gt;i'm trying to add a contact form into a page template with contact form 7. however, i want the destination email ...</td>\n",
       "      <td>php  wordpress  advanced-custom-fields  shortcode  contact-form-7</td>\n",
       "      <td>trying cont form page template cont form however want destination email come custom field created code pulls custom ...</td>\n",
       "      <td>add custom field inside shortcode</td>\n",
       "      <td>[trying, cont, form, page, template, cont, form, however, want, destination, email, come, custom, field, created, co...</td>\n",
       "      <td>[add, custom, field, inside, shortcode]</td>\n",
       "      <td>[try, cont, form, page, template, cont, form, however, want, destination, email, come, custom, field, create, code, ...</td>\n",
       "      <td>[add, custom, field, inside, shortcode]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46753</th>\n",
       "      <td>(some function) is not defined with sympy lambdify</td>\n",
       "      <td>&lt;p&gt;so i'm writing a script that evaluates taylor series. however, i want it to evaluate for all types of functions. ...</td>\n",
       "      <td>python  numpy  sympy</td>\n",
       "      <td>writing script evaluates taylor series however want evaluate types functions tried example using function x x sy sym...</td>\n",
       "      <td>function defined sympy lambdify</td>\n",
       "      <td>[writing, script, evaluates, taylor, series, however, want, evaluate, types, functions, tried, example, using, funct...</td>\n",
       "      <td>[function, defined, sympy, lambdify]</td>\n",
       "      <td>[write, script, evaluates, taylor, series, however, want, evaluate, type, function, try, example, use, function, sy,...</td>\n",
       "      <td>[function, define, sympy, lambdify]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48067</th>\n",
       "      <td>how can i pass parameters to a jquery $.getjson callback method?</td>\n",
       "      <td>&lt;p&gt;i'm trying to use jquery to call some custom api via &lt;code&gt;ajax/$.getjson&lt;/code&gt;. &lt;/p&gt;\\n\\n&lt;p&gt;i'm trying to pass a...</td>\n",
       "      <td>javascript  jquery  callback</td>\n",
       "      <td>trying use jquery custom api via ajax getjson trying pass custom value ajax method value getting passed getting over...</td>\n",
       "      <td>pass parameters jquery getjson callback method</td>\n",
       "      <td>[trying, use, jquery, custom, api, via, ajax, getjson, trying, pass, custom, value, ajax, method, value, getting, pa...</td>\n",
       "      <td>[pass, parameters, jquery, getjson, callback, method]</td>\n",
       "      <td>[try, use, jquery, custom, api, via, ajax, getjson, try, pas, custom, value, ajax, method, value, get, pass, get, ov...</td>\n",
       "      <td>[pas, parameter, jquery, getjson, callback, method]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                  Title  \\\n",
       "14013                            add custom field inside a do_shortcode   \n",
       "46753                (some function) is not defined with sympy lambdify   \n",
       "48067  how can i pass parameters to a jquery $.getjson callback method?   \n",
       "\n",
       "                                                                                                                          Body  \\\n",
       "14013  <p>i'm trying to add a contact form into a page template with contact form 7. however, i want the destination email ...   \n",
       "46753  <p>so i'm writing a script that evaluates taylor series. however, i want it to evaluate for all types of functions. ...   \n",
       "48067  <p>i'm trying to use jquery to call some custom api via <code>ajax/$.getjson</code>. </p>\\n\\n<p>i'm trying to pass a...   \n",
       "\n",
       "                                                                      Tags  \\\n",
       "14013   php  wordpress  advanced-custom-fields  shortcode  contact-form-7    \n",
       "46753                                                python  numpy  sympy    \n",
       "48067                                        javascript  jquery  callback    \n",
       "\n",
       "                                                                                                                  FilteredBody  \\\n",
       "14013  trying cont form page template cont form however want destination email come custom field created code pulls custom ...   \n",
       "46753  writing script evaluates taylor series however want evaluate types functions tried example using function x x sy sym...   \n",
       "48067  trying use jquery custom api via ajax getjson trying pass custom value ajax method value getting passed getting over...   \n",
       "\n",
       "                                         FilteredTitle  \\\n",
       "14013                add custom field inside shortcode   \n",
       "46753                  function defined sympy lambdify   \n",
       "48067  pass parameters jquery getjson callback method    \n",
       "\n",
       "                                                                                                                    BodyTokens  \\\n",
       "14013  [trying, cont, form, page, template, cont, form, however, want, destination, email, come, custom, field, created, co...   \n",
       "46753  [writing, script, evaluates, taylor, series, however, want, evaluate, types, functions, tried, example, using, funct...   \n",
       "48067  [trying, use, jquery, custom, api, via, ajax, getjson, trying, pass, custom, value, ajax, method, value, getting, pa...   \n",
       "\n",
       "                                                 TitleTokens  \\\n",
       "14013                [add, custom, field, inside, shortcode]   \n",
       "46753                   [function, defined, sympy, lambdify]   \n",
       "48067  [pass, parameters, jquery, getjson, callback, method]   \n",
       "\n",
       "                                                                                                                    BodyLemmas  \\\n",
       "14013  [try, cont, form, page, template, cont, form, however, want, destination, email, come, custom, field, create, code, ...   \n",
       "46753  [write, script, evaluates, taylor, series, however, want, evaluate, type, function, try, example, use, function, sy,...   \n",
       "48067  [try, use, jquery, custom, api, via, ajax, getjson, try, pas, custom, value, ajax, method, value, get, pass, get, ov...   \n",
       "\n",
       "                                               TitleLemmas  \n",
       "14013              [add, custom, field, inside, shortcode]  \n",
       "46753                  [function, define, sympy, lambdify]  \n",
       "48067  [pas, parameter, jquery, getjson, callback, method]  "
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfCleaned.sample(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11a28466",
   "metadata": {
    "id": "11a28466"
   },
   "source": [
    "# SAUVEGARDE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "8d1831ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "dfCleaned.drop(['FilteredBody','FilteredTitle','BodyTokens','TitleTokens'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "2d03615c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dfCleaned5000.jbl.bz2']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(dfCleaned.sample(5000),'dfCleaned5000.jbl.bz2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "631630f8",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "631630f8",
    "outputId": "9dc04ef2-ba2e-459b-9de7-66e2bb3f31fa"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['dfCleaned.jbl.bz2']"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "joblib.dump(dfCleaned,'dfCleaned.jbl.bz2')\n",
    "\n",
    "#joblib.dump(dfCleaned,\"drive/MyDrive/Colab Notebooks/Datasets/dfCleaned.jbl.bz2\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0dd91f20",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0dd91f20",
    "outputId": "789ca6da-adc8-49c5-8a54-6978e2bd3a24",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "678.3514068126678"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "elapsed=time.time() - start\n",
    "elapsed"
   ]
  }
 ],
 "metadata": {
  "accelerator": "TPU",
  "colab": {
   "collapsed_sections": [
    "3961885b"
   ],
   "name": "P06_01_Categorisez-automatiquement-des-questions_Nettoyage.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3.8",
   "language": "python",
   "name": "python3.8"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}