Quora_Question_Similarity.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "MRP-fAQedMTd"
   },
   "source": [
    "<h2> 3.6 Featurizing text data with tfidf weighted word-vectors </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "collapsed": false,
    "id": "-3IbomL8dMTi",
    "outputId": "3fa8eb7c-ddf2-4f98-edee-0c49db6502e8"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "import time\n",
    "import warnings\n",
    "import numpy as np\n",
    "from nltk.corpus import stopwords\n",
    "from sklearn.preprocessing import normalize\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "import sys\n",
    "import os \n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "# exctract word2vec vectors\n",
    "# https://github.com/explosion/spaCy/issues/1721\n",
    "# http://landinghub.visualstudio.com/visual-cpp-build-tools\n",
    "import spacy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "collapsed": true,
    "id": "j5XNgVyLdMT7"
   },
   "outputs": [],
   "source": [
    "# avoid decoding problems\n",
    "df = pd.read_csv(\"train.csv\")\n",
    " \n",
    "# encode questions to unicode\n",
    "# https://stackoverflow.com/a/6812069\n",
    "# ----------------- python 2 ---------------------\n",
    "# df['question1'] = df['question1'].apply(lambda x: unicode(str(x),\"utf-8\"))\n",
    "# df['question2'] = df['question2'].apply(lambda x: unicode(str(x),\"utf-8\"))\n",
    "# ----------------- python 3 ---------------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#prepro_features_train.csv (Simple Preprocessing Feartures)\n",
    "#nlp_features_train.csv (NLP Features)\n",
    "if os.path.isfile('nlp_features_train.csv'):\n",
    "    dfnlp = pd.read_csv(\"nlp_features_train.csv\",encoding='latin-1')\n",
    "else:\n",
    "    print(\"download nlp_features_train.csv from drive or run previous notebook\")\n",
    "\n",
    "if os.path.isfile('df_fe_without_preprocessing_train.csv'):\n",
    "    dfppro = pd.read_csv(\"df_fe_without_preprocessing_train.csv\",encoding='latin-1')\n",
    "else:\n",
    "    print(\"download df_fe_without_preprocessing_train.csv from drive or run previous notebook\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.utils import shuffle\n",
    "import math \n",
    "dfnlp = shuffle(dfnlp, random_state=42)\n",
    "\n",
    "dfnlp_train = dfnlp[0:math.ceil(0.80*(df.shape[0]))]\n",
    "dfnlp_test = dfnlp[(math.ceil(0.80*(df.shape[0]))):df.shape[0]]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dfppro = shuffle(dfppro, random_state=42)\n",
    "\n",
    "\n",
    "dfppro_train = dfppro[0:math.ceil(0.80*(df.shape[0]))]\n",
    "dfppro_test = dfppro[(math.ceil(0.80*(df.shape[0]))):df.shape[0]]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# split the dataset into train and test set after shuffling\n",
    "\n",
    "import math\n",
    "\n",
    "\n",
    "df = shuffle(df, random_state=42)\n",
    "\n",
    "# split the dataset into train and test\n",
    "\n",
    "df_train = df[0:math.ceil(0.80*(df.shape[0]))]\n",
    "\n",
    "df_test = df[(math.ceil(0.80*(df.shape[0]))):df.shape[0]]\n",
    "\n",
    "\n",
    "\n",
    "df_train['question1'] = df_train['question1'].apply(lambda x: str(x))\n",
    "df_train['question2'] = df_train['question2'].apply(lambda x: str(x))\n",
    "\n",
    "\n",
    "df_test['question1'] = df_test['question1'].apply(lambda x: str(x))\n",
    "df_test['question2'] = df_test['question2'].apply(lambda x: str(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": false
   },
   "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>qid1</th>\n",
       "      <th>qid2</th>\n",
       "      <th>question1</th>\n",
       "      <th>question2</th>\n",
       "      <th>is_duplicate</th>\n",
       "      <th>freq_qid1</th>\n",
       "      <th>freq_qid2</th>\n",
       "      <th>q1len</th>\n",
       "      <th>q2len</th>\n",
       "      <th>q1_n_words</th>\n",
       "      <th>q2_n_words</th>\n",
       "      <th>word_Common</th>\n",
       "      <th>word_Total</th>\n",
       "      <th>word_share</th>\n",
       "      <th>freq_q1+q2</th>\n",
       "      <th>freq_q1-q2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8067</th>\n",
       "      <td>8067</td>\n",
       "      <td>15738</td>\n",
       "      <td>15739</td>\n",
       "      <td>How do I play Pokémon GO in Korea?</td>\n",
       "      <td>How do I play Pokémon GO in China?</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>34</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.437500</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>368101</th>\n",
       "      <td>368101</td>\n",
       "      <td>12736</td>\n",
       "      <td>104117</td>\n",
       "      <td>What are some of the best side dishes for crab...</td>\n",
       "      <td>What are some good side dishes for buffalo chi...</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>53</td>\n",
       "      <td>51</td>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "      <td>6.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.300000</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70497</th>\n",
       "      <td>70497</td>\n",
       "      <td>121486</td>\n",
       "      <td>121487</td>\n",
       "      <td>Which is more advisable and better material fo...</td>\n",
       "      <td>What is the best server setup for buddypress?</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>96</td>\n",
       "      <td>45</td>\n",
       "      <td>16</td>\n",
       "      <td>8</td>\n",
       "      <td>2.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.083333</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226567</th>\n",
       "      <td>226567</td>\n",
       "      <td>254474</td>\n",
       "      <td>258192</td>\n",
       "      <td>How do I improve logical programming skills?</td>\n",
       "      <td>How can I improve my logical skills for progra...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>44</td>\n",
       "      <td>52</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73186</th>\n",
       "      <td>73186</td>\n",
       "      <td>48103</td>\n",
       "      <td>3062</td>\n",
       "      <td>How close we are to see 3rd world war?</td>\n",
       "      <td>How close is a World War III?</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>31</td>\n",
       "      <td>38</td>\n",
       "      <td>29</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>35</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            id    qid1    qid2  \\\n",
       "8067      8067   15738   15739   \n",
       "368101  368101   12736  104117   \n",
       "70497    70497  121486  121487   \n",
       "226567  226567  254474  258192   \n",
       "73186    73186   48103    3062   \n",
       "\n",
       "                                                question1  \\\n",
       "8067                   How do I play Pokémon GO in Korea?   \n",
       "368101  What are some of the best side dishes for crab...   \n",
       "70497   Which is more advisable and better material fo...   \n",
       "226567       How do I improve logical programming skills?   \n",
       "73186              How close we are to see 3rd world war?   \n",
       "\n",
       "                                                question2  is_duplicate  \\\n",
       "8067                   How do I play Pokémon GO in China?             0   \n",
       "368101  What are some good side dishes for buffalo chi...             0   \n",
       "70497       What is the best server setup for buddypress?             0   \n",
       "226567  How can I improve my logical skills for progra...             1   \n",
       "73186                       How close is a World War III?             1   \n",
       "\n",
       "        freq_qid1  freq_qid2  q1len  q2len  q1_n_words  q2_n_words  \\\n",
       "8067            2          2     34     34           8           8   \n",
       "368101          5          2     53     51          11           9   \n",
       "70497           1          1     96     45          16           8   \n",
       "226567          2          2     44     52           7           9   \n",
       "73186           4         31     38     29           9           7   \n",
       "\n",
       "        word_Common  word_Total  word_share  freq_q1+q2  freq_q1-q2  \n",
       "8067            7.0        16.0    0.437500           4           0  \n",
       "368101          6.0        20.0    0.300000           7           3  \n",
       "70497           2.0        24.0    0.083333           2           0  \n",
       "226567          4.0        16.0    0.250000           4           0  \n",
       "73186           3.0        16.0    0.187500          35          27  "
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfppro_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": false
   },
   "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>qid1</th>\n",
       "      <th>qid2</th>\n",
       "      <th>question1</th>\n",
       "      <th>question2</th>\n",
       "      <th>is_duplicate</th>\n",
       "      <th>cwc_min</th>\n",
       "      <th>cwc_max</th>\n",
       "      <th>csc_min</th>\n",
       "      <th>csc_max</th>\n",
       "      <th>...</th>\n",
       "      <th>ctc_max</th>\n",
       "      <th>last_word_eq</th>\n",
       "      <th>first_word_eq</th>\n",
       "      <th>abs_len_diff</th>\n",
       "      <th>mean_len</th>\n",
       "      <th>token_set_ratio</th>\n",
       "      <th>token_sort_ratio</th>\n",
       "      <th>fuzz_ratio</th>\n",
       "      <th>fuzz_partial_ratio</th>\n",
       "      <th>longest_substr_ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8067</th>\n",
       "      <td>8067</td>\n",
       "      <td>15738</td>\n",
       "      <td>15739</td>\n",
       "      <td>how do i play pokémon go in korea</td>\n",
       "      <td>how do i play pokémon go in china</td>\n",
       "      <td>0</td>\n",
       "      <td>0.749981</td>\n",
       "      <td>0.749981</td>\n",
       "      <td>0.999975</td>\n",
       "      <td>0.999975</td>\n",
       "      <td>...</td>\n",
       "      <td>0.874989</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>90</td>\n",
       "      <td>81</td>\n",
       "      <td>88</td>\n",
       "      <td>88</td>\n",
       "      <td>0.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>368101</th>\n",
       "      <td>368101</td>\n",
       "      <td>12736</td>\n",
       "      <td>104117</td>\n",
       "      <td>what are some of the best side dishes for crab...</td>\n",
       "      <td>what are some good side dishes for buffalo chi...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.399992</td>\n",
       "      <td>0.399992</td>\n",
       "      <td>0.999975</td>\n",
       "      <td>0.666656</td>\n",
       "      <td>...</td>\n",
       "      <td>0.545450</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>73</td>\n",
       "      <td>0.326923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70497</th>\n",
       "      <td>70497</td>\n",
       "      <td>121486</td>\n",
       "      <td>121487</td>\n",
       "      <td>which is more advisable and better material fo...</td>\n",
       "      <td>what is the best server setup for buddypress</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.499988</td>\n",
       "      <td>0.249997</td>\n",
       "      <td>...</td>\n",
       "      <td>0.124999</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>41</td>\n",
       "      <td>39</td>\n",
       "      <td>39</td>\n",
       "      <td>44</td>\n",
       "      <td>0.108696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226567</th>\n",
       "      <td>226567</td>\n",
       "      <td>254474</td>\n",
       "      <td>258192</td>\n",
       "      <td>how do i improve logical programming skills</td>\n",
       "      <td>how can i improve my logical skills for progra...</td>\n",
       "      <td>1</td>\n",
       "      <td>0.999975</td>\n",
       "      <td>0.999975</td>\n",
       "      <td>0.666644</td>\n",
       "      <td>0.399992</td>\n",
       "      <td>...</td>\n",
       "      <td>0.666659</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>96</td>\n",
       "      <td>89</td>\n",
       "      <td>72</td>\n",
       "      <td>68</td>\n",
       "      <td>0.288889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73186</th>\n",
       "      <td>73186</td>\n",
       "      <td>48103</td>\n",
       "      <td>3062</td>\n",
       "      <td>how close we are to see 3rd world war</td>\n",
       "      <td>how close is a world war iii</td>\n",
       "      <td>1</td>\n",
       "      <td>0.749981</td>\n",
       "      <td>0.599988</td>\n",
       "      <td>0.333322</td>\n",
       "      <td>0.249994</td>\n",
       "      <td>...</td>\n",
       "      <td>0.444440</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>81</td>\n",
       "      <td>71</td>\n",
       "      <td>68</td>\n",
       "      <td>62</td>\n",
       "      <td>0.366667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            id    qid1    qid2  \\\n",
       "8067      8067   15738   15739   \n",
       "368101  368101   12736  104117   \n",
       "70497    70497  121486  121487   \n",
       "226567  226567  254474  258192   \n",
       "73186    73186   48103    3062   \n",
       "\n",
       "                                                question1  \\\n",
       "8067                   how do i play pokémon go in korea    \n",
       "368101  what are some of the best side dishes for crab...   \n",
       "70497   which is more advisable and better material fo...   \n",
       "226567       how do i improve logical programming skills    \n",
       "73186              how close we are to see 3rd world war    \n",
       "\n",
       "                                                question2  is_duplicate  \\\n",
       "8067                   how do i play pokémon go in china              0   \n",
       "368101  what are some good side dishes for buffalo chi...             0   \n",
       "70497       what is the best server setup for buddypress              0   \n",
       "226567  how can i improve my logical skills for progra...             1   \n",
       "73186                       how close is a world war iii              1   \n",
       "\n",
       "         cwc_min   cwc_max   csc_min   csc_max          ...           \\\n",
       "8067    0.749981  0.749981  0.999975  0.999975          ...            \n",
       "368101  0.399992  0.399992  0.999975  0.666656          ...            \n",
       "70497   0.000000  0.000000  0.499988  0.249997          ...            \n",
       "226567  0.999975  0.999975  0.666644  0.399992          ...            \n",
       "73186   0.749981  0.599988  0.333322  0.249994          ...            \n",
       "\n",
       "         ctc_max  last_word_eq  first_word_eq  abs_len_diff  mean_len  \\\n",
       "8067    0.874989           0.0            1.0           0.0       8.0   \n",
       "368101  0.545450           0.0            1.0           2.0      10.0   \n",
       "70497   0.124999           0.0            0.0           8.0      12.0   \n",
       "226567  0.666659           0.0            1.0           2.0       8.0   \n",
       "73186   0.444440           0.0            1.0           2.0       8.0   \n",
       "\n",
       "        token_set_ratio  token_sort_ratio  fuzz_ratio  fuzz_partial_ratio  \\\n",
       "8067                 90                81          88                  88   \n",
       "368101               73                73          73                  73   \n",
       "70497                41                39          39                  44   \n",
       "226567               96                89          72                  68   \n",
       "73186                81                71          68                  62   \n",
       "\n",
       "        longest_substr_ratio  \n",
       "8067                0.800000  \n",
       "368101              0.326923  \n",
       "70497               0.108696  \n",
       "226567              0.288889  \n",
       "73186               0.366667  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfnlp_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "collapsed": false,
    "id": "RU3HqJXwdMUj"
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "# merge texts\n",
    "questions = list(df_train['question1']) + list(df_train['question2'])\n",
    "\n",
    "tfidf = TfidfVectorizer(lowercase=False, )\n",
    "tfidf.fit_transform(questions)\n",
    "\n",
    "question1_tfidf = tfidf.transform(df_train['question1'])\n",
    "question2_tfidf = tfidf.transform(df_train['question2'])\n",
    "\n",
    "# dict key:word and value:tf-idf score\n",
    "word2tfidf = dict(zip(tfidf.get_feature_names(), tfidf.idf_))\n",
    "\n",
    "\n",
    "\n",
    "question1_tfidf_test = tfidf.transform(df_test['question1'])\n",
    "question2_tfidf_test = tfidf.transform(df_test['question2'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80858, 98972)"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "question1_tfidf_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(323432, 98972)"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "question1_tfidf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from scipy.sparse import hstack\n",
    "\n",
    "question_features_tfidf= hstack((question1_tfidf, question2_tfidf))\n",
    "question_features_tfidf_test = hstack((question1_tfidf_test, question2_tfidf_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80858, 197944)"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "question_features_tfidf_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "collapsed": true,
    "id": "2JKI2yT4dMUv"
   },
   "source": [
    "- After we find TF-IDF scores, we convert each question to a weighted average of word2vec vectors by these scores.\n",
    "- here we use a pre-trained GLOVE model which comes free with \"Spacy\".  https://spacy.io/usage/vectors-similarity\n",
    "- It is trained on Wikipedia and therefore, it is stronger in terms of word semantics. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df1_train = dfnlp_train.drop(['id','qid1','qid2','question1','question2','is_duplicate'],axis=1)\n",
    "df1_test = dfnlp_test.drop(['id','qid1','qid2','question1','question2','is_duplicate'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df2_train = dfppro_train.drop(['id','qid1','qid2','question1','question2','is_duplicate'],axis=1)\n",
    "df2_test = dfppro_test.drop(['id','qid1','qid2','question1','question2','is_duplicate'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "collapsed": false,
    "id": "Ozz83vh4dMWU",
    "outputId": "e5b30f77-2849-4b08-9949-0912ec0db418",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of features in nlp dataframe : 15\n",
      "Number of features in preprocessed dataframe : 11\n",
      "Number of features in question1 w2v  dataframe : 98972\n",
      "Number of features in question2 w2v  dataframe : 98972\n",
      "Number of features in final dataframe  : 197970\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of features in nlp dataframe :\", df1_train.shape[1])\n",
    "print(\"Number of features in preprocessed dataframe :\", df2_train.shape[1])\n",
    "print(\"Number of features in question1 w2v  dataframe :\", question1_tfidf.shape[1])\n",
    "print(\"Number of features in question2 w2v  dataframe :\", question2_tfidf.shape[1])\n",
    "print(\"Number of features in final dataframe  :\", df1_train.shape[1]+df2_train.shape[1]+question1_tfidf.shape[1]+question2_tfidf.shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del question1_tfidf,question1_tfidf_test,question2_tfidf,question2_tfidf_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# final features in train and test sets\n",
    "tfidf_features_train = hstack((question_features_tfidf,df1_train,df2_train))\n",
    "tfidf_features_test = hstack((question_features_tfidf_test,df1_test,df2_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tfidf_features_train = tfidf_features_train.tocsr()\n",
    "tfidf_features_test = tfidf_features_test.tocsr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_test = dfnlp_test['is_duplicate']\n",
    "y_train = dfnlp_train['is_duplicate']\n",
    "\n",
    "X_train = tfidf_features_train\n",
    "X_test = tfidf_features_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80858, 197970)\n"
     ]
    }
   ],
   "source": [
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import re\n",
    "import time\n",
    "import warnings\n",
    "import sqlite3\n",
    "from sqlalchemy import create_engine # database connection\n",
    "import csv\n",
    "import os\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "import datetime as dt\n",
    "import numpy as np\n",
    "from nltk.corpus import stopwords\n",
    "from sklearn.decomposition import TruncatedSVD\n",
    "from sklearn.preprocessing import normalize\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.manifold import TSNE\n",
    "import seaborn as sns\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics.classification import accuracy_score, log_loss\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from collections import Counter\n",
    "from scipy.sparse import hstack\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "from collections import Counter, defaultdict\n",
    "from sklearn.calibration import CalibratedClassifierCV\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "import math\n",
    "from sklearn.metrics import normalized_mutual_info_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.linear_model import SGDClassifier\n",
    "\n",
    "\n",
    "from sklearn import model_selection\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import precision_recall_curve, auc, roc_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# This function plots the confusion matrices given y_i, y_i_hat.\n",
    "def plot_confusion_matrix(test_y, predict_y):\n",
    "    C = confusion_matrix(test_y, predict_y)\n",
    "    # C = 9,9 matrix, each cell (i,j) represents number of points of class i are predicted class j\n",
    "    \n",
    "    A =(((C.T)/(C.sum(axis=1))).T)\n",
    "    #divid each element of the confusion matrix with the sum of elements in that column\n",
    "    \n",
    "    # C = [[1, 2],\n",
    "    #     [3, 4]]\n",
    "    # C.T = [[1, 3],\n",
    "    #        [2, 4]]\n",
    "    # C.sum(axis = 1)  axis=0 corresonds to columns and axis=1 corresponds to rows in two diamensional array\n",
    "    # C.sum(axix =1) = [[3, 7]]\n",
    "    # ((C.T)/(C.sum(axis=1))) = [[1/3, 3/7]\n",
    "    #                           [2/3, 4/7]]\n",
    "\n",
    "    # ((C.T)/(C.sum(axis=1))).T = [[1/3, 2/3]\n",
    "    #                           [3/7, 4/7]]\n",
    "    # sum of row elements = 1\n",
    "    \n",
    "    B =(C/C.sum(axis=0))\n",
    "    #divid each element of the confusion matrix with the sum of elements in that row\n",
    "    # C = [[1, 2],\n",
    "    #     [3, 4]]\n",
    "    # C.sum(axis = 0)  axis=0 corresonds to columns and axis=1 corresponds to rows in two diamensional array\n",
    "    # C.sum(axix =0) = [[4, 6]]\n",
    "    # (C/C.sum(axis=0)) = [[1/4, 2/6],\n",
    "    #                      [3/4, 4/6]] \n",
    "    plt.figure(figsize=(20,4))\n",
    "    \n",
    "    labels = [1,2]\n",
    "    # representing A in heatmap format\n",
    "    cmap=sns.light_palette(\"blue\")\n",
    "    plt.subplot(1, 3, 1)\n",
    "    sns.heatmap(C, annot=True, cmap=cmap, fmt=\".3f\", xticklabels=labels, yticklabels=labels)\n",
    "    plt.xlabel('Predicted Class')\n",
    "    plt.ylabel('Original Class')\n",
    "    plt.title(\"Confusion matrix\")\n",
    "    \n",
    "    plt.subplot(1, 3, 2)\n",
    "    sns.heatmap(B, annot=True, cmap=cmap, fmt=\".3f\", xticklabels=labels, yticklabels=labels)\n",
    "    plt.xlabel('Predicted Class')\n",
    "    plt.ylabel('Original Class')\n",
    "    plt.title(\"Precision matrix\")\n",
    "    \n",
    "    plt.subplot(1, 3, 3)\n",
    "    # representing B in heatmap format\n",
    "    sns.heatmap(A, annot=True, cmap=cmap, fmt=\".3f\", xticklabels=labels, yticklabels=labels)\n",
    "    plt.xlabel('Predicted Class')\n",
    "    plt.ylabel('Original Class')\n",
    "    plt.title(\"Recall matrix\")\n",
    "    \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------- Distribution of output variable in train data ----------\n",
      "Class 0:  0.6309981696307106 Class 1:  0.36900183036928935\n",
      "---------- Distribution of output variable in train data ----------\n",
      "Class 0:  0.36998194365430753 Class 1:  0.36998194365430753\n"
     ]
    }
   ],
   "source": [
    "print(\"-\"*10, \"Distribution of output variable in train data\", \"-\"*10)\n",
    "train_distr = Counter(y_train)\n",
    "train_len = len(y_train)\n",
    "print(\"Class 0: \",int(train_distr[0])/train_len,\"Class 1: \", int(train_distr[1])/train_len)\n",
    "print(\"-\"*10, \"Distribution of output variable in train data\", \"-\"*10)\n",
    "test_distr = Counter(y_test)\n",
    "test_len = len(y_test)\n",
    "print(\"Class 0: \",int(test_distr[1])/test_len, \"Class 1: \",int(test_distr[1])/test_len)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Random Model </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log loss on Test Data using Random Model 0.8841844551414406\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIwAAAEWCAYAAAAEkwwtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSZJ6LBhJkiRJkiSp5/8BY9OT/HzoZ9EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x288 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# we need to generate 9 numbers and the sum of numbers should be 1\n",
    "# one solution is to genarate 9 numbers and divide each of the numbers by their sum\n",
    "# ref: https://stackoverflow.com/a/18662466/4084039\n",
    "# we create a output array that has exactly same size as the CV data\n",
    "test_len = len(y_test)\n",
    "predicted_y = np.zeros((test_len,2))\n",
    "for i in range(test_len):\n",
    "    rand_probs = np.random.rand(1,2)\n",
    "    predicted_y[i] = ((rand_probs/sum(sum(rand_probs)))[0])\n",
    "print(\"Log loss on Test Data using Random Model\",log_loss(y_test, predicted_y, eps=1e-15))\n",
    "\n",
    "predicted_y =np.argmax(predicted_y, axis=1)\n",
    "plot_confusion_matrix(y_test, predicted_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>Logistic Regression </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For values of alpha =  1e-05 The log loss is: 0.485077280985135\n",
      "For values of alpha =  0.0001 The log loss is: 0.47985347171947945\n",
      "For values of alpha =  0.001 The log loss is: 0.47453870685316213\n",
      "For values of alpha =  0.01 The log loss is: 0.46557300167694415\n",
      "For values of alpha =  0.1 The log loss is: 0.46698806572637674\n",
      "For values of alpha =  1 The log loss is: 0.4950783211129927\n",
      "For values of alpha =  10 The log loss is: 0.5473762777532921\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbAAAAEWCAYAAAAHC8LZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For values of best alpha =  0.01 The train log loss is: 0.4641414328019522\n",
      "For values of best alpha =  0.01 The test log loss is: 0.46557300167694415\n",
      "Total number of data points : 80858\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIAAAAEWCAYAAAAer+yjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1440x288 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "alpha = [10 ** x for x in range(-5, 2)] # hyperparam for SGD classifier.\n",
    "\n",
    "# read more about SGDClassifier() at http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.SGDClassifier.html\n",
    "# ------------------------------\n",
    "# default parameters\n",
    "# SGDClassifier(loss=’hinge’, penalty=’l2’, alpha=0.0001, l1_ratio=0.15, fit_intercept=True, max_iter=None, tol=None, \n",
    "# shuffle=True, verbose=0, epsilon=0.1, n_jobs=1, random_state=None, learning_rate=’optimal’, eta0=0.0, power_t=0.5, \n",
    "# class_weight=None, warm_start=False, average=False, n_iter=None)\n",
    "\n",
    "# some of methods\n",
    "# fit(X, y[, coef_init, intercept_init, …])\tFit linear model with Stochastic Gradient Descent.\n",
    "# predict(X)\tPredict class labels for samples in X.\n",
    "\n",
    "#-------------------------------\n",
    "# video link: \n",
    "#------------------------------\n",
    "\n",
    "\n",
    "log_error_array=[]\n",
    "for i in alpha:\n",
    "    clf = SGDClassifier(alpha=i, penalty='l2', loss='log', random_state=42)\n",
    "    clf.fit(X_train, y_train)\n",
    "    sig_clf = CalibratedClassifierCV(clf, method=\"sigmoid\")\n",
    "    sig_clf.fit(X_train, y_train)\n",
    "    predict_y = sig_clf.predict_proba(X_test)\n",
    "    log_error_array.append(log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "    print('For values of alpha = ', i, \"The log loss is:\",log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(alpha, log_error_array,c='g')\n",
    "for i, txt in enumerate(np.round(log_error_array,3)):\n",
    "    ax.annotate((alpha[i],np.round(txt,3)), (alpha[i],log_error_array[i]))\n",
    "plt.grid()\n",
    "plt.title(\"Cross Validation Error for each alpha\")\n",
    "plt.xlabel(\"Alpha i's\")\n",
    "plt.ylabel(\"Error measure\")\n",
    "plt.show()\n",
    "\n",
    "\n",
    "best_alpha = np.argmin(log_error_array)\n",
    "clf = SGDClassifier(alpha=alpha[best_alpha], penalty='l2', loss='log', random_state=42)\n",
    "clf.fit(X_train, y_train)\n",
    "sig_clf = CalibratedClassifierCV(clf, method=\"sigmoid\")\n",
    "sig_clf.fit(X_train, y_train)\n",
    "\n",
    "predict_y = sig_clf.predict_proba(X_train)\n",
    "print('For values of best alpha = ', alpha[best_alpha], \"The train log loss is:\",log_loss(y_train, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "predict_y = sig_clf.predict_proba(X_test)\n",
    "print('For values of best alpha = ', alpha[best_alpha], \"The test log loss is:\",log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "predicted_y =np.argmax(predict_y,axis=1)\n",
    "print(\"Total number of data points :\", len(predicted_y))\n",
    "plot_confusion_matrix(y_test, predicted_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>Linear SVM with TFIDF </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For values of alpha =  1e-05 The log loss is: 0.4583939704579279\n",
      "For values of alpha =  0.0001 The log loss is: 0.47545291322132216\n",
      "For values of alpha =  0.001 The log loss is: 0.507625736716709\n",
      "For values of alpha =  0.01 The log loss is: 0.5208397731487039\n",
      "For values of alpha =  0.1 The log loss is: 0.506146856572133\n",
      "For values of alpha =  1 The log loss is: 0.5895302967410011\n",
      "For values of alpha =  10 The log loss is: 0.6034962960972355\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbAAAAEWCAYAAAAHC8LZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For values of best alpha =  1e-05 The train log loss is: 0.45750234733998535\n",
      "For values of best alpha =  1e-05 The test log loss is: 0.4583939704579279\n",
      "Total number of data points : 80858\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIAAAAEWCAYAAAAer+yjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 1440x288 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "alpha = [10 ** x for x in range(-5, 2)] # hyperparam for SGD classifier.\n",
    "\n",
    "# read more about SGDClassifier() at http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.SGDClassifier.html\n",
    "# ------------------------------\n",
    "# default parameters\n",
    "# SGDClassifier(loss=’hinge’, penalty=’l2’, alpha=0.0001, l1_ratio=0.15, fit_intercept=True, max_iter=None, tol=None, \n",
    "# shuffle=True, verbose=0, epsilon=0.1, n_jobs=1, random_state=None, learning_rate=’optimal’, eta0=0.0, power_t=0.5, \n",
    "# class_weight=None, warm_start=False, average=False, n_iter=None)\n",
    "\n",
    "# some of methods\n",
    "# fit(X, y[, coef_init, intercept_init, …])\tFit linear model with Stochastic Gradient Descent.\n",
    "# predict(X)\tPredict class labels for samples in X.\n",
    "\n",
    "#-------------------------------\n",
    "# video link: \n",
    "#------------------------------\n",
    "\n",
    "\n",
    "log_error_array=[]\n",
    "for i in alpha:\n",
    "    clf = SGDClassifier(alpha=i, penalty='l1', loss='hinge', random_state=42)\n",
    "    clf.fit(X_train, y_train)\n",
    "    sig_clf = CalibratedClassifierCV(clf, method=\"sigmoid\")\n",
    "    sig_clf.fit(X_train, y_train)\n",
    "    predict_y = sig_clf.predict_proba(X_test)\n",
    "    log_error_array.append(log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "    print('For values of alpha = ', i, \"The log loss is:\",log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.plot(alpha, log_error_array,c='g')\n",
    "for i, txt in enumerate(np.round(log_error_array,3)):\n",
    "    ax.annotate((alpha[i],np.round(txt,3)), (alpha[i],log_error_array[i]))\n",
    "plt.grid()\n",
    "plt.title(\"Cross Validation Error for each alpha\")\n",
    "plt.xlabel(\"Alpha i's\")\n",
    "plt.ylabel(\"Error measure\")\n",
    "plt.show()\n",
    "\n",
    "\n",
    "best_alpha = np.argmin(log_error_array)\n",
    "clf = SGDClassifier(alpha=alpha[best_alpha], penalty='l1', loss='hinge', random_state=42)\n",
    "clf.fit(X_train, y_train)\n",
    "sig_clf = CalibratedClassifierCV(clf, method=\"sigmoid\")\n",
    "sig_clf.fit(X_train, y_train)\n",
    "\n",
    "predict_y = sig_clf.predict_proba(X_train)\n",
    "print('For values of best alpha = ', alpha[best_alpha], \"The train log loss is:\",log_loss(y_train, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "predict_y = sig_clf.predict_proba(X_test)\n",
    "print('For values of best alpha = ', alpha[best_alpha], \"The test log loss is:\",log_loss(y_test, predict_y, labels=clf.classes_, eps=1e-15))\n",
    "predicted_y =np.argmax(predict_y,axis=1)\n",
    "print(\"Total number of data points :\", len(predicted_y))\n",
    "plot_confusion_matrix(y_test, predicted_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Conclusion </h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. TFIDF vectorizer is used to obtain the features to find the duplicate questions between quora similarity pairs.\n",
    "2. Total number of features used are: 197970\n",
    "3. Whole dataset is used to split the samples into training and testing. \n",
    "4. 80% of the samples are used for training and the remaining samples are used for testing.\n",
    "5. From the results below, it can be observed that TFIDF features worked better than TFIDF W2V while \n",
    "training logistic regression and linear svm modesl."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Results </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-----------------+---------------------+----------+-----------+--------+------------+\n",
      "| Vectorizer Name |   Classifier Name   | Log loss | Precision | Recall | F1-measure |\n",
      "+-----------------+---------------------+----------+-----------+--------+------------+\n",
      "|      TFIDF      |     Random Model    |  0.8841  |   0.630   | 0.500  |   0.557    |\n",
      "|      TFIDF      | Logistic Regression |  0.4655  |   0.772   | 0.896  |   0.829    |\n",
      "|      TFIDF      |      Linear SVM     |  0.4583  |   0.783   | 0.881  |   0.829    |\n",
      "+-----------------+---------------------+----------+-----------+--------+------------+\n"
     ]
    }
   ],
   "source": [
    "# THE RESULTS ARE HARDCODED AS THE EXPERIMENTS ARE RUN ON DIFFERENT TIMES.\n",
    "# PLEASE CONSIDER\n",
    "\n",
    "\n",
    "from prettytable import PrettyTable\n",
    "t = PrettyTable(['Vectorizer Name', 'Classifier Name','Log loss','Precision', 'Recall', 'F1-measure'])\n",
    "t.add_row(['TFIDF', 'Random Model', \"0.8841\", \"0.630\",\"0.500\",\"0.557\"])\n",
    "t.add_row(['TFIDF', 'Logistic Regression',\"0.4655\",\"0.772\",\"0.896\",\"0.829\"])\n",
    "t.add_row(['TFIDF', 'Linear SVM',\"0.4583\",\"0.783\",\"0.881\",\"0.829\"])\n",
    "print(t)"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "colab": {
   "collapsed_sections": [],
   "name": "3.Q_Mean_W2V.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}