myExploration.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('dark_background')\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import csv\n",
    "\n",
    "\n",
    "from tqdm import tqdm\n",
    "from functools import partial\n",
    "from scipy import sparse\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer, CountVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "colors = [\"#2965CC\", \"#29A634\", \"#D99E0B\", \"#D13913\", \"#8F398F\",\n",
    "          \"#00B3A4\", \"#DB2C6F\", \"#9BBF30\", \"#96622D\", \"#7157D9\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Quelques outils pour visualiser les données et les résultats de prédiction.\n",
    "On cherche notamment à :\n",
    "* observer la distribution des catégories de l'ensemble d'entrainement\n",
    "* observer la performance prédictive pour chaque catégorie"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chargement des données"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataFolder = './challenge_data'\n",
    "xPath = os.path.join(dataFolder, 'input_train.csv')\n",
    "yPath = os.path.join(dataFolder,\n",
    "                     'output_train.csv')\n",
    "\n",
    "# Loading, parsing and spliting training and testing data\n",
    "x = pd.read_csv(xPath, delimiter=';', usecols=[0, 1]).values\n",
    "y = pd.read_csv(yPath, delimiter=';', usecols=[0, 1]).values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Global Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "51\n"
     ]
    }
   ],
   "source": [
    "SPELL_CORRECTION = False\n",
    "NUMBER_CATEGORIES = len(set(y[:,1]))\n",
    "print(NUMBER_CATEGORIES)\n",
    "STOP_WORDS = ['alors','au','aucuns','aussi','autre','avant','avec','avoir','bon','car',\n",
    "              'ce','cela','ces','ceux','chaque','ci','comme','comment','dans','des',\n",
    "              'du','dedans','dehors','depuis','devrait','doit','donc','dos','début','elle',\n",
    "              'elles','en','encore','essai','est','et','eu','fait','faites','fois',\n",
    "              'font','hors','ici','il','ils','je','juste','la','le','les',\n",
    "              'leur','là','ma','maintenant','mais','mes','mine','moins','mon','mot',\n",
    "              'même','ni','nommés','notre','nous','ou','où','par','parce','pas',\n",
    "              'peut','peu','plupart','pour','pourquoi','quand','que','quel','quelle','quelles',\n",
    "              'quels','qui','sa','sans','ses','seulement','si','sien','son','sont',\n",
    "              'sous','soyez','sujet','sur','ta','tandis','tellement','tels','tes','ton',\n",
    "              'tous','tout','trop','très','tu','voient','vont','votre','vous','vu',\n",
    "              'ça','étaient','état','étions','été','être', 'de', 'un', 'une', 'ai', 'ne', 'on', 'me', ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spelling correction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6467\n"
     ]
    }
   ],
   "source": [
    "if SPELL_CORRECTION:\n",
    "    correctionsDict = {}\n",
    "    correctionsPath = 'challenge_data/corrections.csv'\n",
    "\n",
    "    with open(correctionsPath, 'r') as f:\n",
    "        for key, val in csv.reader(f):\n",
    "            correctionsDict[key] = val\n",
    "\n",
    "    def spellingCorrection(sentence, correctionsDict={}, verbose=False):\n",
    "        correctedSentence = []\n",
    "        nb_corrections = 0\n",
    "        for word in sentence.split():\n",
    "            if word in correctionsDict.keys():\n",
    "                word = correctionsDict[word]\n",
    "                nb_corrections += 1\n",
    "            correctedSentence.append(word)\n",
    "        if verbose:\n",
    "            return ' '.join(correctedSentence), nb_corrections\n",
    "        else:\n",
    "            return ' '.join(correctedSentence)\n",
    "\n",
    "    nb_corr = 0\n",
    "    for i, s in enumerate(x[:,1]):\n",
    "        cor = spellingCorrection(s, correctionsDict, verbose = True)\n",
    "        x[i] = cor[0]\n",
    "        nb_corr += cor[1]\n",
    "print(nb_corr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Distribution du vocabulaire"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def parseWord(wordCount, totalWordCount, density, item):\n",
    "    if density:\n",
    "        return [item[0], wordCount[0, item[1]] / totalWordCount]\n",
    "    else:\n",
    "        return [item[0], wordCount[0, item[1]]]\n",
    "\n",
    "def vectorizeVocabulary(corpus, verbose=False, density=False):\n",
    "    # Generate word tokens\n",
    "    countVectorizer = CountVectorizer(input='content')\n",
    "    countVector = countVectorizer.fit_transform(corpus)\n",
    "    vocabulary = countVectorizer.vocabulary_\n",
    "    wordCount = np.sum(countVector, axis=0)\n",
    "    totalWordCount = np.sum(wordCount)\n",
    "\n",
    "    vocabulary = list(map(partial(parseWord, wordCount,\n",
    "                                  totalWordCount, density), vocabulary.items()))\n",
    "    \n",
    "    # Sort words by usage\n",
    "    sortedVocabulary = sorted(vocabulary, key=lambda x: x[1], reverse=True)\n",
    "    \n",
    "    if verbose:\n",
    "        print(\"countVector.shape: {}\".format(str(countVector.shape)))\n",
    "        print(\"wordCount.shape: {}\".format(str(wordCount.shape)))\n",
    "        print(sortedVocabulary[:5])\n",
    "    \n",
    "    return sortedVocabulary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotVocabulary(vocabulary, numberWords=50):\n",
    "    labels = [word[0] for word in vocabulary[:numberWords]]\n",
    "    values = [word[1] for word in vocabulary[:numberWords]]\n",
    "    yPosition = np.arange(len(labels))\n",
    "\n",
    "    plt.figure(figsize=(15,8))\n",
    "    plt.bar(yPosition, values, color=colors[0])\n",
    "    plt.xticks(yPosition, labels, rotation=90)\n",
    "    plt.ylabel('Unique count')\n",
    "    plt.title('Word')\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4gAAAH/CAYAAAAVCPOeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGiASBEEQBEEQBEEQAID/Ahba9fECM5v4AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f651521d9e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9870\n"
     ]
    }
   ],
   "source": [
    "sentences = x[:,1]\n",
    "sortedVocabulary = vectorizeVocabulary(sentences, verbose=False, density=True)\n",
    "plotVocabulary(sortedVocabulary)\n",
    "print(len(sortedVocabulary))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Distribution des catégories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "countCategories = [0] * NUMBER_CATEGORIES\n",
    "\n",
    "for point in y:\n",
    "    countCategories[point[1]] += 1\n",
    "\n",
    "indexedCountCategories = list(zip(list(range(NUMBER_CATEGORIES)), countCategories))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4UAAAHiCAYAAABfvmlrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/R9YGItvbCMbOQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe84a3b3390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "yPosition = np.arange(NUMBER_CATEGORIES)\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "plt.bar(yPosition, countCategories, color=colors[0])\n",
    "\n",
    "plt.ylabel('Unique count')\n",
    "plt.title('Categories')\n",
    "axes = plt.gca()\n",
    "#axes.set_ylim([0,50])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(4, 76),\n",
       " (9, 54),\n",
       " (10, 68),\n",
       " (12, 89),\n",
       " (24, 64),\n",
       " (26, 72),\n",
       " (29, 77),\n",
       " (30, 52),\n",
       " (43, 91),\n",
       " (47, 82)]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selectedCategories = list(filter(lambda x:  50 < x[1] <= 100, indexedCountCategories))\n",
    "selectedCategories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4UAAAHlCAYAAABCu1nTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSiEAAECFKYUAAAAVphQCAABU2P8HG62rku6dZw0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe84a328e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of categories with morethan 100 points: 20\n",
      "Mean of number of points in categories with more than 100 points: 337.0\n"
     ]
    }
   ],
   "source": [
    "selectedCategories = list(filter(lambda x: x[1] > 100, indexedCountCategories))\n",
    "\n",
    "labels = [category[0] for category in selectedCategories]\n",
    "values = [category[1] for category in selectedCategories]\n",
    "yPosition = np.arange(len(selectedCategories))\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "plt.bar(yPosition, values, color=colors[0])\n",
    "plt.xticks(yPosition, labels, rotation=90)\n",
    "plt.ylabel('Unique count')\n",
    "plt.title('Word')\n",
    " \n",
    "plt.show()\n",
    "\n",
    "print('Number of categories with more'\n",
    "      'than 100 points: {}'.format(str(len(selectedCategories))))\n",
    "print('Mean of number of points in categories with more than 100 points: {}'.format(np.mean([cat[1] for cat in selectedCategories])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Distribution du vocabulaire par catégorie"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "countVector.shape: (104, 822)\n",
      "wordCount.shape: (1, 822)\n",
      "[['de', 0.046953046953046952], ['je', 0.033300033300033303], ['et', 0.02763902763902764], ['ai', 0.022311022311022312], ['que', 0.02197802197802198]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA4IAAAIUCAYAAABPQZKLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe84a25ecf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "category = 27\n",
    "selectedSentences = x[y == category][1:]\n",
    "sortedVocabularyCategory =\\\n",
    "    vectorizeVocabulary(list(map(lambda x: str(x), selectedSentences)),\n",
    "                        verbose=True, density=True)\n",
    "plotVocabulary(sortedVocabularyCategory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmsAAAJOCAYAAADoLUBhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe84a1acc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "vocabularyDict = dict(sortedVocabulary)\n",
    "plt.figure(figsize=(10,10))\n",
    "m = 0\n",
    "selectedWords = {}\n",
    "\n",
    "for category in range(NUMBER_CATEGORIES):\n",
    "    selectedSentences = x[y == category][1:]\n",
    "    sortedVocabularyCategory = vectorizeVocabulary(list(map(lambda x: str(x),\n",
    "                                                            selectedSentences)),\n",
    "                                                   density=True)\n",
    "\n",
    "    for word in sortedVocabularyCategory[:20]:\n",
    "        if not word[0] in STOP_WORDS:\n",
    "            m = max(word[1], m)\n",
    "            plt.scatter(word[1], vocabularyDict[word[0]],\n",
    "                        color=colors[category % len(colors)], label=str(category))\n",
    "            if category in selectedWords:\n",
    "                selectedWords[category].append([*word, vocabularyDict[word[0]]])\n",
    "            else:\n",
    "                selectedWords[category] = [[*word, vocabularyDict[word[0]]]]\n",
    "\n",
    "plt.plot([0, 1], color=colors[0])\n",
    "plt.ylim(0, 1.1 * m)\n",
    "plt.xlim(0, 1.1 * m)\n",
    "plt.ylabel('Unique count')\n",
    "plt.title('Word')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Explore the medicaments names "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Liste de médicaments regroupant les libéllés ATC et lesdénominations de spécialité, de taille: 8390\n",
      "Sample of medicament names:  ['a 313 200  pour cent', 'a 313 50 000 u.i', 'abacavir', 'abacavir/lamivudine', 'abacavir/lamivudine pharma', 'abacavir/lamivudine pharos', 'abamipharm', 'abboticine', 'abelcet', 'abstral']\n"
     ]
    }
   ],
   "source": [
    "MEDICAMENTS = []\n",
    "\n",
    "with open(os.path.join(dataFolder, 'medicList.txt')) as f:\n",
    "    for line in f:\n",
    "        MEDICAMENTS.append(line.lower().rstrip())\n",
    "\n",
    "print('Liste de médicaments regroupant les libéllés ATC et les'\n",
    "      'dénominations de spécialité, de taille: {}'.format(len(MEDICAMENTS)))\n",
    "print('Sample of medicament names: ', MEDICAMENTS[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8028/8028 [00:14<00:00, 551.04it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9404 medicaments occurences in the dataset\n",
      "600 unique medicament names occuring in the dataset\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "medicaments_occurences = []\n",
    "nb_occurences = 0\n",
    "unique_occurences = []\n",
    "for xx in tqdm(x):\n",
    "    occur = []\n",
    "    for w in xx[1].split():\n",
    "        w = w.lower()\n",
    "        if w in MEDICAMENTS:\n",
    "            occur.append(w)\n",
    "            nb_occurences +=1\n",
    "            if w not in unique_occurences:\n",
    "                unique_occurences.append(w)\n",
    "    medicaments_occurences.append(occur)\n",
    "print('{} medicaments occurences in the dataset'.format(nb_occurences))\n",
    "print('{} unique medicament names occuring in the dataset'.format(len(unique_occurences)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Length of the sentences "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max length: 701 / min length: 1 / mean length:17.4 / median length 10.0 / 99 percentile:112.0\n",
      "number of  sentences of length > 100 : 105\n",
      "number of  sentences of length > 200 : 18\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtUAAADSCAYAAACFKTz6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f65147b20b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sent_lengths = [len(xx.split()) for xx in x[:,1]]\n",
    "plt.figure(figsize=(12,3))\n",
    "plt.hist(sent_lengths, bins = 80)\n",
    "plt.xlim((0, 200))\n",
    "plt.title('distribution if sentences lengths')\n",
    "print('max length: {} / min length: {} / mean length:{} / median length {} / 99 percentile:{}'.format(\n",
    "    max(sent_lengths), min(sent_lengths), \n",
    "    round(np.mean(sent_lengths),1), np.median(sent_lengths), np.percentile(sent_lengths, 99)))\n",
    "print('number of  sentences of length > 100 : {}'.format(np.sum([1 for e in sent_lengths if e >100])))\n",
    "print('number of  sentences of length > 200 : {}'.format(np.sum([1 for e in sent_lengths if e >200])))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def selectCat(x, y, cat_index, medicaments_occurences = medicaments_occurences):\n",
    "    cat_ix = []\n",
    "    for xx, yy in zip(x, y):\n",
    "        if yy[1] == cat_index:\n",
    "            cat_ix.append(xx[0])\n",
    "    cat_ix = np.array(cat_ix)\n",
    "    selected_questions = x[cat_ix]\n",
    "    selected_medics = [medicaments_occurences[ix] for ix in cat_ix]\n",
    "    return selected_questions, selected_medics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# try to see if there is a correletation between medicament names and the categories\n",
    "meds_by_category = {}\n",
    "for cat in set(y[:,1]):\n",
    "    meds_by_category[cat] = []\n",
    "    \n",
    "for yy, meds in zip(y, medicaments_occurences):\n",
    "    for med in meds:\n",
    "        if med not in meds_by_category[yy[1]]:\n",
    "            meds_by_category[yy[1]].append(med)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA34AAAHsCAYAAACNGH91AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncDrOogAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe849dd6550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "labels = [nb_cat for nb_cat in np.arange(NUMBER_CATEGORIES)]\n",
    "values = [len(meds_cat) for meds_cat in list(meds_by_category.values())]\n",
    "ratio_values = [v/l for (v, l) in zip(values, countCategories)]\n",
    "yPosition = np.arange(NUMBER_CATEGORIES)\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "plt.title('Divesity of medicament names in the different categories')\n",
    "plt.subplot(211)\n",
    "plt.bar(yPosition, values, color=colors[0])\n",
    "plt.xticks(yPosition, labels, rotation=90)\n",
    "plt.ylabel('Unique medicament occurences count')\n",
    "\n",
    "plt.subplot(212)\n",
    "plt.bar(yPosition, ratio_values, color=colors[0])\n",
    "plt.xticks(yPosition, labels, rotation=90)\n",
    "plt.ylabel('medicament occurences ratio (wrt size of the category)')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Explore the variety of questions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 45 questions in category 46\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[56,\n",
       "        'le prix du générique lisinopril 20mg est passé subitement de 12,39€ à 13,25€ , est ce normal?'],\n",
       "       [75, 'je souhaiterais savoir quel est le prix du vermifuge humain ?'],\n",
       "       [457,\n",
       "        'je voudrais commencer une cure de cystine b6. j ai entendu dire qu en france le prix etait de 3 € ?'],\n",
       "       [484,\n",
       "        \"j'ai arrêté cycléane 20 il y a 4 ans et là j'attends mes règles pour pouvoir la reprendre. je voudrais savoir combien coûtent les 3 plaquettes maintenant ?\"],\n",
       "       [703,\n",
       "        'obligé de payer mon rivotril 50 euros la bouteille , normal ?'],\n",
       "       [1003, 'cérazette ? Combien ça coûte ?'],\n",
       "       [1047,\n",
       "        'obligé de payer mon rivotril 50 euros la bouteille , normal ?'],\n",
       "       [1091, 'tarif trinordiol?'],\n",
       "       [1602, 'Quel est le prix du générique biogaran de mercilon ?'],\n",
       "       [1718, 'prix cycléane 20 ?']], dtype=object)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_index = 46\n",
    "cat_questions, med_questions = selectCat(x, y, cat_index)\n",
    "print('There are {} questions in category {}'.format(len(cat_questions), cat_index))\n",
    "cat_questions[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### See short questions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "min_length = 15\n",
    "short_questions = []\n",
    "for xx in x:\n",
    "    if len(xx[1]) <= min_length:\n",
    "        short_questions.append((xx[1], y[xx[0],1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('urbanyl?', 22),\n",
       " ('deroxat ??', 22),\n",
       " ('C quoi neisvac?', 22),\n",
       " ('REJET', 14),\n",
       " ('neoral', 22),\n",
       " ('norset', 22),\n",
       " ('optilova 20 ?', 22),\n",
       " ('la damiana?', 22),\n",
       " ('méningitec ?', 22),\n",
       " ('exacyl ??', 22)]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "short_questions[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extract important vocabulary for each category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_top_k(x, y, k, stop_words = STOP_WORDS):\n",
    "    best_vocab = []\n",
    "    nb_cat = len(set(y[:,1]))\n",
    "    for cat in np.arange(nb_cat):\n",
    "        questions = selectCat(x, y, cat)[0][:,1]\n",
    "        cat_vocab = vectorizeVocabulary(corpus=questions)\n",
    "        cat_best_vocab = []\n",
    "        for w in cat_vocab:\n",
    "            if len(cat_best_vocab) < k:\n",
    "                if w[0] not in stop_words:\n",
    "                    cat_best_vocab.append(w)           \n",
    "        best_vocab.append(cat_best_vocab)\n",
    "    return best_vocab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_vocab = extract_top_k(x, y, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['faire', 102],\n",
       " ['dois', 46],\n",
       " ['plus', 46],\n",
       " ['me', 35],\n",
       " ['suis', 34],\n",
       " ['jours', 26],\n",
       " ['mal', 24],\n",
       " ['mois', 22],\n",
       " ['pris', 21],\n",
       " ['prise', 21]]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ix = 11\n",
    "best_vocab[ix]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "195 questions in class 11\n",
      "[[34\n",
      "  \"je viens de commencer le traitement cystine b6 depuis 3 jours et j'ai des gaz que puis-je faire ?\"]\n",
      " [38\n",
      "  \"le medecin spécialiste gastro m'a préscrit lansoprazole. mon médecin traitant m'a préscit olmetec 20 mg pour hypertension, et depuis qiue je les prends ensemble une réaction se forme. des petits points rouges très demengeant et ambetant sont apparus sur ma quisse droite. j'ai arreté l'olmetec , mais les points ne sont pas encore partis après 3 jours.. que faire ou prendre ?\"]\n",
      " [79 'help! dépendance ixprim?']\n",
      " [111\n",
      "  \"j'ai utilisé l'elavil, le naxproxen a petites doses pour des problèmes de dos. avec le celebrex c'était beaucoup de spasme dans le haut du dos, entre les côtes. le problème de dos à dégénéré en problèmes d'anxiété et je prends en ce moment du rivotril au soir et au couché et du desyrel aussi au coucheé. toujours ses courbatures, bien sûr elle passe facilement au levée, je crois que c'est les médicaments, mais j'en ai pas la certitude. Que faire?\"]\n",
      " [237\n",
      "  \"j'ai pris minesse pour la 1ere fois mais au bout de 15 jours, j'ai toujours mes règles!! cela veut dire que cette pilule ne me convient pas ou dois-je encore attendre 1 mois ou 2 avant de le savoir?\"]\n",
      " [285 \"J'ai de l'eczema suite à la prise d'extranaze, que dois-je faire ?\"]\n",
      " [306 'help! dépendance ixprim?']\n",
      " [314\n",
      "  \"depuis que j'ai pris la pillule jasminelle y a quelques mois j'avais des petites taches de sang ou marron sans gravité. aujourd'hui j'ai des saignements comme le début des règles assez léger. je ne sais pas ce qu'il faut que je fasse.\"]\n",
      " [329\n",
      "  \"je suis passée sous diu cuivre, j'ai maintenant des plaques rouges de boutons sur les deux joues. Que faire?\"]\n",
      " [450\n",
      "  \"Mon gygy m'a donné une pillule daily gé que j'ai prise 3 jours apres le retrait du sterilet mirena. arrivée a la fin de ma plaquette, pas de regles durant l'arret et j'ai repris normalement ma deuxieme plaquette, mais rien ni regles, ni ovulation. que dois-je faire ?\"]\n",
      " [459 'qlaira et sécheresse vaginale, changer pour un sterilet?']\n",
      " [506\n",
      "  'j ai mis mon 3eme patch evra du mois hier et depuis 3j j ai des perte marront foncé. je voudrai aussi savoir ce qui pourrais reduire les effet du patch ?']\n",
      " [526 'Je fais une réaction après avoir pris repevax. Que dois je faire?']\n",
      " [596\n",
      "  'j ai commence optimisette en novembre 2015 mais depuis que j ai commence la 2eme boite en fevrier soit la 4eme plaquette c est l horreur! j ai des frissons toute la journee. que faire ?']\n",
      " [628\n",
      "  \"après 2 comprimés de minesse, voilà que depuis ce matin j'ai une sorte d'engourdissement de la jambe gauche.. que faire ?\"]\n",
      " [692\n",
      "  \"8 jrs avant la fin de ma plaquette d'evepar j'ai eu des saignement et des douleurs au bas ventre et au bas du dos dois je m'inquiéter car la douleur est plus importante que d'habitude ?\"]\n",
      " [701\n",
      "  \"j'ai fait le vaccin gerdazil (le premier) il y a environ un moi , mais depuis hier jai un sorte de plaque sous la peau tres douloureuze a l'endroit ou j'ai fait le vaccin . quelqu'un pourrais me dire si je doit consulter?\"]\n",
      " [838\n",
      "  'Mes règles sont très longues depuis le retrait de mirena et la pose de ut380, que dois-je faire ?']\n",
      " [1079 'Le buspirone me fatigue, que faire ?']\n",
      " [1088\n",
      "  \"j'ai pris mon dernier comprime de luteran et depuis avant hier j'ai tres mal au seins, ces derniers gonflent pas mal , nausee et tout.  comment les diminuer ?\"]]\n"
     ]
    }
   ],
   "source": [
    "questions = selectCat(x, y, ix)[0]\n",
    "print('{} questions in class {}'.format(len(questions),ix))\n",
    "print(questions[:20])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}