Logistic Regression - Amazon food reviews data set.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Amazon Fine Food Reviews Analysis\n",
    "\n",
    "\n",
    "Data Source: https://www.kaggle.com/snap/amazon-fine-food-reviews <br>\n",
    "\n",
    "EDA: https://nycdatascience.com/blog/student-works/amazon-fine-foods-visualization/\n",
    "\n",
    "\n",
    "The Amazon Fine Food Reviews dataset consists of reviews of fine foods from Amazon.<br>\n",
    "\n",
    "Number of reviews: 568,454<br>\n",
    "Number of users: 256,059<br>\n",
    "Number of products: 74,258<br>\n",
    "Timespan: Oct 1999 - Oct 2012<br>\n",
    "Number of Attributes/Columns in data: 10 \n",
    "\n",
    "Attribute Information:\n",
    "\n",
    "1. Id\n",
    "2. ProductId - unique identifier for the product\n",
    "3. UserId - unqiue identifier for the user\n",
    "4. ProfileName\n",
    "5. HelpfulnessNumerator - number of users who found the review helpful\n",
    "6. HelpfulnessDenominator - number of users who indicated whether they found the review helpful or not\n",
    "7. Score - rating between 1 and 5\n",
    "8. Time - timestamp for the review\n",
    "9. Summary - brief summary of the review\n",
    "10. Text - text of the review\n",
    "\n",
    "\n",
    "#### Objective:\n",
    "Given a review, determine whether the review is positive (Rating of 4 or 5) or negative (rating of 1 or 2).\n",
    "\n",
    "<br>\n",
    "[Q] How to determine if a review is positive or negative?<br>\n",
    "<br> \n",
    "[Ans] We could use the Score/Rating. A rating of 4 or 5 could be cosnidered a positive review. A review of 1 or 2 could be considered negative. A review of 3 is nuetral and ignored. This is an approximate and proxy way of determining the polarity (positivity/negativity) of a review.\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the data\n",
    "\n",
    "The dataset is available in two forms\n",
    "1. .csv file\n",
    "2. SQLite Database\n",
    "\n",
    "In order to load the data, We have used the SQLITE dataset as it easier to query the data and visualise the data efficiently.\n",
    "<br> \n",
    "\n",
    "Here as we only want to get the global sentiment of the recommendations (positive or negative), we will purposefully ignore all Scores equal to 3. If the score id above 3, then the recommendation wil be set to \"positive\". Otherwise, it will be set to \"negative\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\gensim\\utils.py:1197: UserWarning: detected Windows; aliasing chunkize to chunkize_serial\n",
      "  warnings.warn(\"detected Windows; aliasing chunkize to chunkize_serial\")\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "import os\n",
    "import re\n",
    "import sqlite3\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import nltk\n",
    "import string\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sb\n",
    "import pickle\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import roc_curve,auc\n",
    "\n",
    "from nltk.stem.porter import PorterStemmer\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.stem.wordnet import WordNetLemmatizer\n",
    "from nltk.stem import PorterStemmer\n",
    "\n",
    "from gensim.models import Word2Vec\n",
    "from gensim.models import KeyedVectors\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler \n",
    "#TSNE\n",
    "from sklearn.manifold import TSNE\n",
    "from bs4 import BeautifulSoup\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Temporarily Suppressing Warnings\n",
    "def fxn():\n",
    "    warnings.warn(\"deprecated\", DeprecationWarning)\n",
    "\n",
    "with warnings.catch_warnings():\n",
    "    warnings.simplefilter(\"ignore\")\n",
    "    fxn()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [1]. Reading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of data point in our data (100000, 10)\n"
     ]
    },
    {
     "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>ProductId</th>\n",
       "      <th>UserId</th>\n",
       "      <th>ProfileName</th>\n",
       "      <th>HelpfulnessNumerator</th>\n",
       "      <th>HelpfulnessDenominator</th>\n",
       "      <th>Score</th>\n",
       "      <th>Time</th>\n",
       "      <th>Summary</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>B001E4KFG0</td>\n",
       "      <td>A3SGXH7AUHU8GW</td>\n",
       "      <td>delmartian</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1303862400</td>\n",
       "      <td>Good Quality Dog Food</td>\n",
       "      <td>I have bought several of the Vitality canned d...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>B00813GRG4</td>\n",
       "      <td>A1D87F6ZCVE5NK</td>\n",
       "      <td>dll pa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1346976000</td>\n",
       "      <td>Not as Advertised</td>\n",
       "      <td>Product arrived labeled as Jumbo Salted Peanut...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>B000LQOCH0</td>\n",
       "      <td>ABXLMWJIXXAIN</td>\n",
       "      <td>Natalia Corres \"Natalia Corres\"</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1219017600</td>\n",
       "      <td>\"Delight\" says it all</td>\n",
       "      <td>This is a confection that has been around a fe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>B000UA0QIQ</td>\n",
       "      <td>A395BORC6FGVXV</td>\n",
       "      <td>Karl</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1307923200</td>\n",
       "      <td>Cough Medicine</td>\n",
       "      <td>If you are looking for the secret ingredient i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>B006K2ZZ7K</td>\n",
       "      <td>A1UQRSCLF8GW1T</td>\n",
       "      <td>Michael D. Bigham \"M. Wassir\"</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1350777600</td>\n",
       "      <td>Great taffy</td>\n",
       "      <td>Great taffy at a great price.  There was a wid...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Id   ProductId          UserId                      ProfileName  \\\n",
       "0   1  B001E4KFG0  A3SGXH7AUHU8GW                       delmartian   \n",
       "1   2  B00813GRG4  A1D87F6ZCVE5NK                           dll pa   \n",
       "2   3  B000LQOCH0   ABXLMWJIXXAIN  Natalia Corres \"Natalia Corres\"   \n",
       "3   4  B000UA0QIQ  A395BORC6FGVXV                             Karl   \n",
       "4   5  B006K2ZZ7K  A1UQRSCLF8GW1T    Michael D. Bigham \"M. Wassir\"   \n",
       "\n",
       "   HelpfulnessNumerator  HelpfulnessDenominator  Score        Time  \\\n",
       "0                     1                       1      1  1303862400   \n",
       "1                     0                       0      0  1346976000   \n",
       "2                     1                       1      1  1219017600   \n",
       "3                     3                       3      0  1307923200   \n",
       "4                     0                       0      1  1350777600   \n",
       "\n",
       "                 Summary                                               Text  \n",
       "0  Good Quality Dog Food  I have bought several of the Vitality canned d...  \n",
       "1      Not as Advertised  Product arrived labeled as Jumbo Salted Peanut...  \n",
       "2  \"Delight\" says it all  This is a confection that has been around a fe...  \n",
       "3         Cough Medicine  If you are looking for the secret ingredient i...  \n",
       "4            Great taffy  Great taffy at a great price.  There was a wid...  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# using the SQLite Table to read data.\n",
    "# con = sqlite3.connect('./amazon-fine-food-reviews/database.sqlite')\n",
    "\n",
    "con = sqlite3.connect('C:/Users/Saraswathi/Music/Appliedai/Data/amazon-fine-food-reviews/database.sqlite')\n",
    "\n",
    "#filetering only positve and negative reviews\n",
    "#reviews not taking in to consideration with score = 3\n",
    "\n",
    "filtered_data = pd.read_sql_query(\"\"\" SELECT * FROM Reviews WHERE Score != 3 LIMIT 100000\"\"\", con) \n",
    "\n",
    "# Give reviews with Score>3 a positive rating, and reviews with a score<3 a negative rating.\n",
    "def partition( x ):\n",
    "    if x > 3:\n",
    "        return 1 #positive\n",
    "#         return 1\n",
    "    else:\n",
    "        return 0 #negative\n",
    "\n",
    "#changing reviews with score less than 3 to be positive and vice versa\n",
    "actual_score = filtered_data['Score']\n",
    "positivenegative = actual_score.map(partition)\n",
    "filtered_data['Score']=positivenegative\n",
    "print('Number of data point in our data',filtered_data.shape)\n",
    "filtered_data.head(5)    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  Exploratory Data Analysis\n",
    "\n",
    "## [2] Data Cleaning: Deduplication\n",
    "It is observed (as shown in the table below) that the reviews data had many duplicate entries. Hence it was necessary to remove duplicates in order to get unbiased results for the analysis of the data.  Following is an example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "display = pd.read_sql_query(\"\"\"\n",
    "SELECT * FROM Reviews\n",
    "WHERE Score != 3 AND UserId=\"AR5J8UI46CURR\"\n",
    "ORDER BY ProductID\n",
    "\"\"\",con)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Id</th>\n",
       "      <th>ProductId</th>\n",
       "      <th>UserId</th>\n",
       "      <th>ProfileName</th>\n",
       "      <th>HelpfulnessNumerator</th>\n",
       "      <th>HelpfulnessDenominator</th>\n",
       "      <th>Score</th>\n",
       "      <th>Time</th>\n",
       "      <th>Summary</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>78445</td>\n",
       "      <td>B000HDL1RQ</td>\n",
       "      <td>AR5J8UI46CURR</td>\n",
       "      <td>Geetha Krishnan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1199577600</td>\n",
       "      <td>LOACKER QUADRATINI VANILLA WAFERS</td>\n",
       "      <td>DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>138317</td>\n",
       "      <td>B000HDOPYC</td>\n",
       "      <td>AR5J8UI46CURR</td>\n",
       "      <td>Geetha Krishnan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1199577600</td>\n",
       "      <td>LOACKER QUADRATINI VANILLA WAFERS</td>\n",
       "      <td>DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>138277</td>\n",
       "      <td>B000HDOPYM</td>\n",
       "      <td>AR5J8UI46CURR</td>\n",
       "      <td>Geetha Krishnan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1199577600</td>\n",
       "      <td>LOACKER QUADRATINI VANILLA WAFERS</td>\n",
       "      <td>DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>73791</td>\n",
       "      <td>B000HDOPZG</td>\n",
       "      <td>AR5J8UI46CURR</td>\n",
       "      <td>Geetha Krishnan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1199577600</td>\n",
       "      <td>LOACKER QUADRATINI VANILLA WAFERS</td>\n",
       "      <td>DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>155049</td>\n",
       "      <td>B000PAQ75C</td>\n",
       "      <td>AR5J8UI46CURR</td>\n",
       "      <td>Geetha Krishnan</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1199577600</td>\n",
       "      <td>LOACKER QUADRATINI VANILLA WAFERS</td>\n",
       "      <td>DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Id   ProductId         UserId      ProfileName  HelpfulnessNumerator  \\\n",
       "0   78445  B000HDL1RQ  AR5J8UI46CURR  Geetha Krishnan                     2   \n",
       "1  138317  B000HDOPYC  AR5J8UI46CURR  Geetha Krishnan                     2   \n",
       "2  138277  B000HDOPYM  AR5J8UI46CURR  Geetha Krishnan                     2   \n",
       "3   73791  B000HDOPZG  AR5J8UI46CURR  Geetha Krishnan                     2   \n",
       "4  155049  B000PAQ75C  AR5J8UI46CURR  Geetha Krishnan                     2   \n",
       "\n",
       "   HelpfulnessDenominator  Score        Time  \\\n",
       "0                       2      5  1199577600   \n",
       "1                       2      5  1199577600   \n",
       "2                       2      5  1199577600   \n",
       "3                       2      5  1199577600   \n",
       "4                       2      5  1199577600   \n",
       "\n",
       "                             Summary  \\\n",
       "0  LOACKER QUADRATINI VANILLA WAFERS   \n",
       "1  LOACKER QUADRATINI VANILLA WAFERS   \n",
       "2  LOACKER QUADRATINI VANILLA WAFERS   \n",
       "3  LOACKER QUADRATINI VANILLA WAFERS   \n",
       "4  LOACKER QUADRATINI VANILLA WAFERS   \n",
       "\n",
       "                                                Text  \n",
       "0  DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...  \n",
       "1  DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...  \n",
       "2  DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...  \n",
       "3  DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...  \n",
       "4  DELICIOUS WAFERS. I FIND THAT EUROPEAN WAFERS ...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As can be seen above the same user has multiple reviews of the with the same values for HelpfulnessNumerator, HelpfulnessDenominator, Score, Time, Summary and Text and on doing analysis it was found that \n",
    "\n",
    "ProductId=B000HDOPZG was Loacker Quadratini Vanilla Wafer Cookies, 8.82-Ounce Packages (Pack of 8)\n",
    "\n",
    "ProductId=B000HDL1RQ was Loacker Quadratini Lemon Wafer Cookies, 8.82-Ounce Packages (Pack of 8) and so on\n",
    "\n",
    "It was inferred after analysis that reviews with same parameters other than ProductId belonged to the same product just having different flavour or quantity. Hence in order to reduce redundancy it was decided to eliminate the rows having same parameters.\n",
    "\n",
    "The method used for the same was that we first sort the data according to ProductId and then just keep the first similar product review and delelte the others. for eg. in the above just the review for ProductId=B000HDL1RQ remains. This method ensures that there is only one representative for each product and deduplication without sorting would lead to possibility of different representatives still existing for the same product."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Sorting data according to ProductId in ascending order\n",
    "sorted_data = filtered_data.sort_values('ProductId',axis=0,ascending= True, inplace=False, kind ='quicksort',na_position='last')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(87775, 10)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Duplication of entries\n",
    "final = sorted_data.drop_duplicates(subset={'UserId','ProfileName','Time','Text'}, keep = 'first' , inplace= False)\n",
    "final.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "87.775"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Checking to see how much % of data still remains\n",
    "(final['Id'].size*1.0)/(filtered_data['Id'].size*1.0)*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observation:- It was also seen that in two rows given below the value of HelpfulnessNumerator is greater than HelpfulnessDenominator which is not practically possible hence these two rows too are removed from calcualtions\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>Id</th>\n",
       "      <th>ProductId</th>\n",
       "      <th>UserId</th>\n",
       "      <th>ProfileName</th>\n",
       "      <th>HelpfulnessNumerator</th>\n",
       "      <th>HelpfulnessDenominator</th>\n",
       "      <th>Score</th>\n",
       "      <th>Time</th>\n",
       "      <th>Summary</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>64422</td>\n",
       "      <td>B000MIDROQ</td>\n",
       "      <td>A161DK06JJMCYF</td>\n",
       "      <td>J. E. Stephens \"Jeanne\"</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1224892800</td>\n",
       "      <td>Bought This for My Son at College</td>\n",
       "      <td>My son loves spaghetti so I didn't hesitate or...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44737</td>\n",
       "      <td>B001EQ55RW</td>\n",
       "      <td>A2V0I904FH7ABY</td>\n",
       "      <td>Ram</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1212883200</td>\n",
       "      <td>Pure cocoa taste with crunchy almonds inside</td>\n",
       "      <td>It was almost a 'love at first bite' - the per...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Id   ProductId          UserId              ProfileName  \\\n",
       "0  64422  B000MIDROQ  A161DK06JJMCYF  J. E. Stephens \"Jeanne\"   \n",
       "1  44737  B001EQ55RW  A2V0I904FH7ABY                      Ram   \n",
       "\n",
       "   HelpfulnessNumerator  HelpfulnessDenominator  Score        Time  \\\n",
       "0                     3                       1      5  1224892800   \n",
       "1                     3                       2      4  1212883200   \n",
       "\n",
       "                                        Summary  \\\n",
       "0             Bought This for My Son at College   \n",
       "1  Pure cocoa taste with crunchy almonds inside   \n",
       "\n",
       "                                                Text  \n",
       "0  My son loves spaghetti so I didn't hesitate or...  \n",
       "1  It was almost a 'love at first bite' - the per...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display = pd.read_sql_query(\"\"\"\n",
    "SELECt * \n",
    "FROM Reviews\n",
    "WHERE Score !=3 AND Id=44737 OR Id=64422\n",
    "ORDER BY ProductId\n",
    "\"\"\",con)\n",
    "display.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "final = final[final.HelpfulnessNumerator <= final.HelpfulnessDenominator]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    73592\n",
       "0    14181\n",
       "Name: Score, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final.shape\n",
    "final['Score'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Text Preprocessing.\n",
    "\n",
    "Now that we have finished deduplication our data requires some preprocessing before we go on further with analysis and making the prediction model.\n",
    "\n",
    "Hence in the Preprocessing phase we do the following in the order below:-\n",
    "\n",
    "1. Begin by removing the html tags\n",
    "2. Remove any punctuations or limited set of special characters like , or . or # etc.\n",
    "3. Check if the word is made up of english letters and is not alpha-numeric\n",
    "4. Check to see if the length of the word is greater than 2 (as it was researched that there is no adjective in 2-letters)\n",
    "5. Convert the word to lowercase\n",
    "6. Remove Stopwords\n",
    "7. Finally Snowball Stemming the word (it was obsereved to be better than Porter Stemming)<br>\n",
    "\n",
    "After which we collect the words used to describe positive and negative reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def decontracted(phrase):\n",
    "    # specific\n",
    "    phrase = re.sub(r\"won't\", \"will not\", phrase)\n",
    "    phrase = re.sub(r\"can\\'t\", \"can not\", phrase)\n",
    "\n",
    "    # general\n",
    "    phrase = re.sub(r\"n\\'t\", \" not\", phrase)\n",
    "    phrase = re.sub(r\"\\'re\", \" are\", phrase)\n",
    "    phrase = re.sub(r\"\\'s\", \" is\", phrase)\n",
    "    phrase = re.sub(r\"\\'d\", \" would\", phrase)\n",
    "    phrase = re.sub(r\"\\'ll\", \" will\", phrase)\n",
    "    phrase = re.sub(r\"\\'t\", \" not\", phrase)\n",
    "    phrase = re.sub(r\"\\'ve\", \" have\", phrase)\n",
    "    phrase = re.sub(r\"\\'m\", \" am\", phrase)\n",
    "    return phrase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "stopwords= set(['br', 'the', 'i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', \"you're\", \"you've\",\\\n",
    "            \"you'll\", \"you'd\", 'your', 'yours', 'yourself', 'yourselves', 'he', 'him', 'his', 'himself', \\\n",
    "            'she', \"she's\", 'her', 'hers', 'herself', 'it', \"it's\", 'its', 'itself', 'they', 'them', 'their',\\\n",
    "            'theirs', 'themselves', 'what', 'which', 'who', 'whom', 'this', 'that', \"that'll\", 'these', 'those', \\\n",
    "            'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'having', 'do', 'does', \\\n",
    "            'did', 'doing', 'a', 'an', 'the', 'and', 'but', 'if', 'or', 'because', 'as', 'until', 'while', 'of', \\\n",
    "            'at', 'by', 'for', 'with', 'about', 'against', 'between', 'into', 'through', 'during', 'before', 'after',\\\n",
    "            'above', 'below', 'to', 'from', 'up', 'down', 'in', 'out', 'on', 'off', 'over', 'under', 'again', 'further',\\\n",
    "            'then', 'once', 'here', 'there', 'when', 'where', 'why', 'how', 'all', 'any', 'both', 'each', 'few', 'more',\\\n",
    "            'most', 'other', 'some', 'such', 'only', 'own', 'same', 'so', 'than', 'too', 'very', \\\n",
    "            's', 't', 'can', 'will', 'just', 'don', \"don't\", 'should', \"should've\", 'now', 'd', 'll', 'm', 'o', 're', \\\n",
    "            've', 'y', 'ain', 'aren', \"aren't\", 'couldn', \"couldn't\", 'didn', \"didn't\", 'doesn', \"doesn't\", 'hadn',\\\n",
    "            \"hadn't\", 'hasn', \"hasn't\", 'haven', \"haven't\", 'isn', \"isn't\", 'ma', 'mightn', \"mightn't\", 'mustn',\\\n",
    "            \"mustn't\", 'needn', \"needn't\", 'shan', \"shan't\", 'shouldn', \"shouldn't\", 'wasn', \"wasn't\", 'weren', \"weren't\", \\\n",
    "            'won', \"won't\", 'wouldn', \"wouldn't\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Combining all the above stundents \n",
    "from tqdm import tqdm\n",
    "preprocessed_reviews = []\n",
    "# tqdm is for printing the status bar\n",
    "# for sentance in tqdm(final['Text'].values):\n",
    "for sentance in final['Text'].values:\n",
    "    sentance = re.sub(r\"http\\S+\",\"\",sentance)\n",
    "    sentance = BeautifulSoup(sentance,'lxml').get_text()\n",
    "    sentance = decontracted(sentance)\n",
    "    sentance = re.sub(\"\\S*\\d\\S*\",\"\",sentance).strip()\n",
    "    sentance = re.sub('[^A-Za-z]+',' ',sentance)\n",
    "    sentance = ' '.join(e.lower() for e in sentance.split() if e.lower() not in stopwords)  \n",
    "    preprocessed_reviews.append(sentance.strip())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add pre processed reviews in to final df\n",
    "# final['preprocessed_reviews'] = preprocessed_reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'frenchbull given nylabone chew since weeks old safe strong bite not break large pieces could choke dinosaur chew perfect many places hold bite dylabone product buy'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocessed_reviews[100]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2><font color='red'>[3.2] Preprocess Summary</font></h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\bs4\\__init__.py:273: UserWarning: \"b'...'\" looks like a filename, not markup. You should probably open this file and pass the filehandle into Beautiful Soup.\n",
      "  ' Beautiful Soup.' % markup)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\bs4\\__init__.py:273: UserWarning: \"b'...'\" looks like a filename, not markup. You should probably open this file and pass the filehandle into Beautiful Soup.\n",
      "  ' Beautiful Soup.' % markup)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\bs4\\__init__.py:273: UserWarning: \"b'...'\" looks like a filename, not markup. You should probably open this file and pass the filehandle into Beautiful Soup.\n",
      "  ' Beautiful Soup.' % markup)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\bs4\\__init__.py:273: UserWarning: \"b'...'\" looks like a filename, not markup. You should probably open this file and pass the filehandle into Beautiful Soup.\n",
      "  ' Beautiful Soup.' % markup)\n"
     ]
    }
   ],
   "source": [
    "##preprocessing for review summary also.\n",
    "\n",
    "# Combining all the above stundents \n",
    "from tqdm import tqdm\n",
    "preprocessed_summary = []\n",
    "# tqdm is for printing the status bar\n",
    "# for sentance in tqdm(final['Summary'].values):\n",
    "for sentance in (final['Summary'].values):\n",
    "\n",
    "    sentance = re.sub(r\"http\\S+\",\"\",sentance)\n",
    "    sentance = BeautifulSoup(sentance,'lxml').get_text()\n",
    "    sentance = decontracted(sentance)\n",
    "    sentance = re.sub(\"\\S*\\d\\S*\",\"\",sentance).strip()\n",
    "    sentance = re.sub('[^A-Za-z]+',' ',sentance)\n",
    "    sentance = ' '.join(e.lower() for e in sentance.split() if e.lower() not in stopwords)  \n",
    "    preprocessed_summary.append(sentance.strip())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'frenchbull dog loves nylabones'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preprocessed_summary[100]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Kxx72u1hLPo1"
   },
   "source": [
    "# [5] Assignment 5: Apply Logistic Regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "KDR9We0oLPo2"
   },
   "source": [
    "<ol>\n",
    "    <li><strong>Apply Logistic Regression on these feature sets</strong>\n",
    "        <ul>\n",
    "            <li><font color='red'>SET 1:</font>Review text, preprocessed one converted into vectors using (BOW)</li>\n",
    "            <li><font color='red'>SET 2:</font>Review text, preprocessed one converted into vectors using (TFIDF)</li>\n",
    "            <li><font color='red'>SET 3:</font>Review text, preprocessed one converted into vectors using (AVG W2v)</li>\n",
    "            <li><font color='red'>SET 4:</font>Review text, preprocessed one converted into vectors using (TFIDF W2v)</li>\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Hyper paramter tuning (find best hyper parameters corresponding the algorithm that you choose)</strong>\n",
    "        <ul>\n",
    "    <li>Find the best hyper parameter which will give the maximum <a href='https://www.appliedaicourse.com/course/applied-ai-course-online/lessons/receiver-operating-characteristic-curve-roc-curve-and-auc-1/'>AUC</a> value</li>\n",
    "    <li>Find the best hyper paramter using k-fold cross validation or simple cross validation data</li>\n",
    "    <li>Use gridsearch cv or randomsearch cv or you can also write your own for loops to do this task of hyperparameter tuning</li>          \n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Pertubation Test</strong>\n",
    "        <ul>\n",
    "    <li>Get the weights W after fit your model with the data X i.e Train data.</li>\n",
    "    <li>Add a noise to the X (X' = X + e) and get the new data set X' (if X is a sparse\n",
    "matrix, X.data+=e)</li>\n",
    "    <li>Fit the model again on data X' and get the weights W'</li>\n",
    "    <li>Add a small eps value(to eliminate the divisible by zero error) to W and W’ i.e\n",
    "W=W+10^-6 and W’ = W’+10^-6</li>\n",
    "    <li>Now find the % change between W and W' (| (W-W') / (W) |)*100)</li>\n",
    "    <li>Calculate the 0th, 10th, 20th, 30th, ...100th percentiles, and observe any sudden rise in the values of percentage_change_vector</li>\n",
    "    <li> Ex: consider your 99th percentile is 1.3 and your 100th percentiles are 34.6, there is sudden rise from 1.3 to 34.6, now calculate the 99.1, 99.2, 99.3,..., 100th percentile values and get the proper value after which there is sudden rise the values, assume it is 2.5</li>\n",
    "            <li> Print the feature names whose % change is more than a threshold x(in our example it's 2.5)</li>\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Sparsity</strong>\n",
    "        <ul>\n",
    "    <li>Calculate sparsity on weight vector obtained after using L1 regularization</li>\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br><font color='red'>NOTE: Do sparsity and multicollinearity for any one of the vectorizers. Bow or tf-idf is recommended.</font>\n",
    "    <br>\n",
    "    <br>\n",
    "    <li><strong>Feature importance</strong>\n",
    "        <ul>\n",
    "    <li>Get top 10 important features for both positive and negative classes separately.</li>\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Feature engineering</strong>\n",
    "        <ul>\n",
    "    <li>To increase the performance of your model, you can also experiment with with feature engineering like :</li>\n",
    "            <ul>\n",
    "            <li>Taking length of reviews as another feature.</li>\n",
    "            <li>Considering some features from review summary as well.</li>\n",
    "        </ul>\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Representation of results</strong>\n",
    "        <ul>\n",
    "    <li>You need to plot the performance of model both on train data and cross validation data for each hyper parameter, like shown in the figure.\n",
    "<!--     <img src='train_cv_auc.JPG' width=300px></li> -->\n",
    "    <li>Once after you found the best hyper parameter, you need to train your model with it, and find the AUC on test data and plot the ROC curve on both train and test.\n",
    "<!--     <img src='train_test_auc.JPG' width=300px></li> -->\n",
    "    <li>Along with plotting ROC curve, you need to print the <a href='https://www.appliedaicourse.com/course/applied-ai-course-online/lessons/confusion-matrix-tpr-fpr-fnr-tnr-1/'>confusion matrix</a> with predicted and original labels of test data points. Please visualize your confusion matrices using <a href='https://seaborn.pydata.org/generated/seaborn.heatmap.html'>seaborn heatmaps.\n",
    "<!--     <img src='confusion_matrix.png' width=300px></li> -->\n",
    "        </ul>\n",
    "    </li>\n",
    "    <br>\n",
    "    <li><strong>Conclusion</strong>\n",
    "        <ul>\n",
    "    <li>You need to summarize the results at the end of the notebook, summarize it in the table format. To print out a table please refer to this prettytable library<a href='http://zetcode.com/python/prettytable/'>  link</a> \n",
    "<!--         <img src='summary.JPG' width=400px> -->\n",
    "    </li>\n",
    "        </ul>\n",
    "</ol>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jiYocvSHLPo4"
   },
   "source": [
    "# Applying Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_auc_score,auc\n",
    "\n",
    "# final['cleaned_text'] = preprocessed_reviews\n",
    "# final['cleaned_summary'] = preprocessed_summary\n",
    "# # As data is time series data. So, first sort the data based on time \n",
    "# final_sort_data = final.sort_values('Time',axis = 0, ascending= True, inplace= False, kind= 'quicksort',na_position='last')\n",
    "\n",
    "# x = final_sort_data['cleaned_text']\n",
    "# # y = np.array(final_sort_data['Score'])\n",
    "# y = final_sort_data['Score'].values\n",
    "\n",
    "# #tarin ,cv, test split\n",
    "# x_train, x_test, y_train, y_test = train_test_split(x, y, test_size = 0.3 , random_state = 0)\n",
    "# # x_train, x_cv, y_train, y_cv = train_test_split(x1, y1, test_size = 0.3, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "final['cleaned_text'] = preprocessed_reviews\n",
    "final['cleaned_summary'] = preprocessed_summary\n",
    "\n",
    "#Sort the data based on time\n",
    "final_sort_data = final.sort_values('Time',axis = 0, ascending= True, inplace= False, kind= 'quicksort',na_position='last')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(26331, 12)\n",
      "(18432, 12)\n",
      "(43008, 12)\n"
     ]
    }
   ],
   "source": [
    "#Train,CV,test split\n",
    "final_train_cv_data = final_sort_data[:int((final_sort_data.shape[0]*70)/100)] # slice first 70% points in training set and rest 30% points in test set. \n",
    "final_sort_test_data = final_sort_data[int((final_sort_data.shape[0]*70)/100)+1:]\n",
    "final_sort_train_data = final_train_cv_data[:int((final_train_cv_data.shape[0]*70)/100)] # slice first 70% points in training set and rest 30% points in test\n",
    "final_sort_cv_data = final_train_cv_data[int((final_train_cv_data.shape[0]*70)/100)+1:]\n",
    "\n",
    "# print(final_train_cv_data.shape)\n",
    "print(final_sort_test_data.shape)\n",
    "print(final_sort_cv_data.shape)\n",
    "print(final_sort_train_data.shape)\n",
    "# print(final_sort_train_data.columns)\n",
    "\n",
    "base_x_train = final_sort_train_data['cleaned_text']\n",
    "y_train = np.array(final_sort_train_data['Score'])\n",
    "base_x_cv = final_sort_cv_data['cleaned_text']\n",
    "y_cv = np.array(final_sort_cv_data['Score'])\n",
    "base_x_test = final_sort_test_data['cleaned_text']\n",
    "y_test = np.array(final_sort_test_data['Score'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "LE_XPVZJLPo4"
   },
   "source": [
    "\n",
    "## [5.1] Logistic Regression on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_iciJvj2LPo4"
   },
   "source": [
    "### [5.1.1] Applying Logistic Regression with L1 regularization on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "count_vect = CountVectorizer()\n",
    "x_train = count_vect.fit_transform(base_x_train)\n",
    "x_test = count_vect.transform(base_x_test)\n",
    "x_cv = count_vect.transform(base_x_cv)\n",
    "\n",
    "#Standardize the data\n",
    "scalar = StandardScaler(with_mean= False)\n",
    "x_train = scalar.fit_transform(x_train)\n",
    "x_cv = scalar.transform(x_cv)\n",
    "x_test = scalar.transform(x_test)\n",
    "\n",
    "C = [10**-3,10**-2,10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l1',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5h4NhywcLPo7"
   },
   "source": [
    "#### [5.1.1.1] Calculating sparsity on weight vector obtained using L1 regularization on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-mmdLzUALPo-"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of non zero element in weight vector  3494\n"
     ]
    }
   ],
   "source": [
    "lr = LogisticRegression(penalty='l1',class_weight = 'balanced',C = optimal_c)\n",
    "lr.fit(x_train,y_train)\n",
    "weight = lr.coef_\n",
    "#Sparsity of vector weight=no of zero in weight vector\n",
    "print('Number of non zero element in weight vector ',np.count_nonzero(weight))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "aRHMa52qLPpA"
   },
   "source": [
    "### [5.1.2] Applying Logistic Regression with L2 regularization on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GBbb9lxlLPpB"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "count_vect = CountVectorizer()\n",
    "x_train = count_vect.fit_transform(base_x_train)\n",
    "x_test = count_vect.transform(base_x_test)\n",
    "x_cv = count_vect.transform(base_x_cv)\n",
    "\n",
    "#Standardize the data\n",
    "scalar = StandardScaler(with_mean= False)\n",
    "x_train = scalar.fit_transform(x_train)\n",
    "x_cv = scalar.transform(x_cv)\n",
    "x_test = scalar.transform(x_test)\n",
    "\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l2',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "UaJkugTKLPpE"
   },
   "source": [
    "#### [5.1.2.1] Performing pertubation test (multicollinearity check) on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Find with weight vector using train without noise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf_before = LogisticRegression(C= 1, penalty= 'l2')\n",
    "clf_before.fit(x_train,y_train)\n",
    "w_after= clf_before.coef_\n",
    "x_train.data+=0.001  #this is used for adding noise to data. If you are getting typecase error, use numpy add with casting set to 'unsafe'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf_after = LogisticRegression(C= 1, penalty= 'l2')\n",
    "clf_after.fit(x_train,y_train)\n",
    "w_before= clf_after.coef_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Find with weight vector using train with noise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average difference in weight vectors:  -6.955799258467983e-05\n"
     ]
    }
   ],
   "source": [
    "diff = w_after - w_before\n",
    "print(\"Average difference in weight vectors: \",np.mean(diff))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Obeservation - Differences in the coefficients of the peturbed model and original model are very less so weight vector of the classifier can be considered for the feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = 1)\n",
    "# lr.fit(x_train,y_train)\n",
    "# w_before = lr.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "E1c5B1nELPpF"
   },
   "outputs": [],
   "source": [
    "# epsilon = np.random.normal(loc = 0, scale = 0.0000001,size = x_train.shape)\n",
    "# X_e.data=X_e.data+np.random.normal(loc=0,scale=0.0001,size=X_e.data.shape)\n",
    "# x_train_epsilon = x_train + epsilon\n",
    "\n",
    "# x_train_epsilon = x_train + np.random.normal(loc = 0, scale = 0.0001,size = x_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = 1)\n",
    "# lr.fit(x_train_epsilon,y_train)\n",
    "# w_after = lr.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# diff = w_after - w_before\n",
    "# print(\"Average difference in weight vectors:- \",np.mean(diff))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eB1rF0BQLPpG"
   },
   "source": [
    "### [5.1.3] Feature Importance on BOW,<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JZTP19dqLPpI"
   },
   "source": [
    "#### [5.1.3.1] Top 10 important features of positive class from<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "oGRpCVJgLPpI"
   },
   "outputs": [],
   "source": [
    "lr = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = 0.001)\n",
    "lr.fit(x_train,y_train)\n",
    "fn = count_vect.get_feature_names()\n",
    "w = list(lr.coef_[0])\n",
    "# print(w)\n",
    "pos_coef =[]\n",
    "neg_coef = []\n",
    "pos_words =[]\n",
    "neg_words =[]\n",
    "\n",
    "for i,c in enumerate(w):\n",
    "    if c > 0:\n",
    "        pos_coef.append(c)\n",
    "        pos_words.append(fn[i]) \n",
    "    else:\n",
    "        neg_coef.append(abs(c))\n",
    "        neg_words.append(fn[i])\n",
    "\n",
    "pos_df = pd.DataFrame(columns= ['words', 'coef'])\n",
    "neg_df = pd.DataFrame(columns= ['words', 'coef'])\n",
    "pos_df['words'] = pos_words\n",
    "neg_df['words'] = neg_words\n",
    "pos_df['coef'] = pos_coef\n",
    "neg_df['coef'] = neg_coef\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 10 Important features of positive class -\n",
      "        words      coef\n",
      "0      great  0.374052\n",
      "1       best  0.224588\n",
      "2  delicious  0.217418\n",
      "3       love  0.178129\n",
      "4    perfect  0.132145\n",
      "5       good  0.132019\n",
      "6      loves  0.129792\n",
      "7       easy  0.101703\n",
      "8  excellent  0.093120\n",
      "9     highly  0.087079\n"
     ]
    }
   ],
   "source": [
    "pos_df = pos_df.sort_values('coef',axis = 0,ascending = False).reset_index(drop = True)\n",
    "\n",
    "print('Top 10 Important features of positive class -\\n',pos_df.head(10))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "UJLYBE1gLPpJ"
   },
   "source": [
    "#### [5.1.3.2] Top 10 important features of negative class from<font color='red'> SET 1</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "c9r7Zjx1LPpK"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 10 Important features of negative class -\n",
      "            words      coef\n",
      "0            not  0.262289\n",
      "1   disappointed  0.125496\n",
      "2          worst  0.102344\n",
      "3            bad  0.088504\n",
      "4          money  0.082809\n",
      "5       horrible  0.081595\n",
      "6          awful  0.078224\n",
      "7       terrible  0.077012\n",
      "8        thought  0.074323\n",
      "9  unfortunately  0.067954\n"
     ]
    }
   ],
   "source": [
    "neg_df = neg_df.sort_values('coef',axis = 0,ascending = False).reset_index(drop = True)\n",
    "\n",
    "print('Top 10 Important features of negative class -\\n',neg_df.head(10))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nKn3VXcZLPpN"
   },
   "source": [
    "## [5.2] Logistic Regression on TFIDF,<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "naTBSzm6LPpN"
   },
   "source": [
    "### [5.2.1] Applying Logistic Regression with L1 regularization on TFIDF,<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "\n",
    "tf_idf_vect = TfidfVectorizer(min_df= 10) \n",
    "x_train = tf_idf_vect.fit_transform(base_x_train)\n",
    "x_test = tf_idf_vect.transform(base_x_test)\n",
    "x_cv = tf_idf_vect.transform(base_x_cv)\n",
    "\n",
    "#Standardize the data\n",
    "scalar = StandardScaler(with_mean= False)\n",
    "x_train = scalar.fit_transform(x_train)\n",
    "x_cv = scalar.transform(x_cv)\n",
    "x_test = scalar.transform(x_test)\n",
    "\n",
    "\n",
    "# tuned_parameters = [{'C': [10**-4, 10**-2, 10**0, 10**2, 10**4]}]\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l1',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',1//optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "NdFShHQJLPpV"
   },
   "source": [
    "### [5.2.2] Applying Logistic Regression with L2 regularization on TFIDF,<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GBbb9lxlLPpB"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py:922: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "tf_idf_vect = TfidfVectorizer(min_df= 10) \n",
    "x_train = tf_idf_vect.fit_transform(base_x_train)\n",
    "x_test = tf_idf_vect.transform(base_x_test)\n",
    "x_cv = tf_idf_vect.transform(base_x_cv)\n",
    "\n",
    "#Standardize the data\n",
    "scalar = StandardScaler(with_mean= False)\n",
    "x_train = scalar.fit_transform(x_train)\n",
    "x_cv = scalar.transform(x_cv)\n",
    "x_test = scalar.transform(x_test)\n",
    "\n",
    "\n",
    "C = [10**-3, 10**-2,10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l2',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzt3Xl8FPX9+PHXezcnEK5wacJ9BhISMCqoKIooQiuoqKjUo6htrdrW1m/RqrTV+rOt1dZKa73PcniCilKp94HKJXLWcAjhkCRAgIQcu/v+/TGbIXcCZLMb9v18PPKY2dnPzrxnF+Y985mZ94iqYowxxgB4wh2AMcaYyGFJwRhjjMuSgjHGGJclBWOMMS5LCsYYY1yWFIwxxrgsKRjTgonI+yJybbjjMMcOSwom4onIZhE5KCIHRGSniDwtIm2qtTlFRN4Vkf0iUigir4vI4Gpt2orIX0VkS3BeOcHXnepYropIUbBtxd//hXJdQ0VEegXXJybcsZjIZknBtBTfV9U2QBYwDLit4g0RGQn8B5gHHA/0Br4CPhGRPsE2ccB/gSHAOKAtcApQAJxUz3IzVbVNpb8/1daoto3t4W6AbYNtIoElBdOiqOpOYCFOcqjwJ+BZVf2bqu5X1d2qegewGPhtsM2VQA/gAlVdo6oBVd2lqner6oLDjUNEfisiL4nI8yKyD7i6jmnxwaOR7cG/v4pIfHAeo0UkV0R+LSI7gadqWc7VIvKJiPw9eAS0TkTG1BGTR0TuEJFvRWSXiDwrIu2Cb38YHO4NHvGMPNx1NtHBkoJpUUQkFTgPyAm+boWzx/9iLc3nAmOD42cDb6vqgSYMZyLwEtAeeKGOab8BRuAksUyco5I7Ks2jG9AR6AlcX8dyTgY2Ap2AGcArItKxlnZXB//OBPoAbYCHg++dHhy2Dx7xfNb41TTRxJKCaSleE5H9wFZgF87GEZwNqgfYUctnduBsSAGS62jTkGUisrfS37mV3vtMVV8LHnUcrGPaFcDvg0clecDvgB9UmkcAmKGqpZXmUd0u4K+qWq6qc4D1wIRa2l0BPKCqG4PJ7zZginVLmcNhScG0FJNUNQkYDQzi0MZ+D86G9bhaPnMckB8cL6ijTUOGq2r7Sn8LK723tZb21acdD3xb6fW3wWkV8lS1pIEYtmnVypXV51HfsmKArg3M3xiXJQXToqjqB8DTwP3B10XAZ8DFtTS/BOfkMsAi4FwRad2U4TRi2nacrqEKPYLT6ptHdSkiIvXMo75l+YDvGrkcYywpmBbpr8BYEak42TwduEpEbhaRJBHpICL3ACNxumsAnsPZi39ZRAYFT8omi8jtIjI+hLHOAu4Qkc7BS1/vAp4/zHl0AW4WkVgRuRhIA2o7OT4L+IWI9A5esnsvMEdVfUAezhFVnyNdERMdLCmYFifYN/8scGfw9cfAucCFOOcNvsW5bPU0Vf0m2KYU52TzOuAdYB/wBU431Of1LO6ravcp/PUww70HWAKsBL4GlgWnHY7Pgf44XWF/ACarakEt7Z7ESX4fApuAEuAmAFUtDn72k+C5kRGHGYOJEmIP2TEmconI1cC1qnpauGMx0cGOFIwxxrgsKRhjjHFZ95ExxhiXHSkYY4xxtbg7HTt16qS9evUKdxjGGNOiLF26NF9VOzfUrsUlhV69erFkyZJwh2GMMS2KiHzbcCvrPjLGGFOJJQVjjDEuSwrGGGNclhSMMca4LCkYY4xxhSwpiMiTwUcCrqrjfRGRh4IPT18pIsNDFYsxxpjGCeWRwtM4D0ivy3k4lR/74zyG8J8hjMUYY0wjhOw+BVX9UER61dNkIs7D1hVYLCLtReQ4VT2SRyYaE5FUFVUIqKKAKijOtCrjFW0BDdScHgi29Qec8UBACQSn+wPqtgmoBl/DwXI/Xo8AFTFUmldwWYEa8dVsW+rz4/MrCbFeNy449NSeyqVyDr1XqV0t7d1P1Na+0rwqWlZ5T9WdjxCAgB9PoByPvwzRcjwBHxLwEeM/CCh7i0rpkOhBNAAaQDSAqB8I4NEAnkAZXn8pAU/soTZUtHPaCgrqxxsoJc5XRJn30LOaxPnmEFU0ODy0YkrwFwjOw107RMF5xEVwzTQ4vVK76uMdh09kwPAz6vjX1jTCefNaClUfXZgbnFYjKYjI9QQfat6jR49mCc7Uzx9Qyv0Byv0BfH7FF9xI+QKK36/4VfEHApSUB9sE2/sDyp7ichJiPMENGPhV3Y2cP6AUl/kpLvOTEOtx5+0LzmP73oN0bBNHIBBcZqBiWeAPBNhdVIZHhPhYL/6Aszx/4NDGMqBKwYEyvB4hPsbjLDu4kfW7G9pDG9/cPcV0SUqostENaM0NcMV4qS/Q8JcXsRQvAWLxkUgpiZQRJ+XE4aMNzuOjY/ATI35niJ9k2UeZxhAv5SRQRitKicNHrPhIoph2UkSJxhEnfuIop6d8R562I4YAHgJ4CeCVQ+Mx+EmgzJmOnxgCeMUZdpG9FGs8guLFT5z4w/x9Nb/P2x4Hx3BSkFqm1VqdT1UfBR4FyM7OjroKfv6AUlLubCgLD5ZT6vNzsMzP/lIfApT7nQ3ud/tKSIj1crDMz5bdxbRNjHU23L4A+QdKAfB6PO7GfMvuYjq2jsMf0ODG19nwbs4vonNSvLMh9wUoDyh5+0tJiPUQUPD5AwTC/Cu0ivPiFcHjEbwVfyJ4BPKLyujaNp74mMptcMdbxXnJP1BKYtsEEmO8eIKf83okOO7Mz+MRBh/Xln0l5XRtmxD8PIgcWpZUtBXwiLPPuO9gOV2S4kEEwZku4vyDl+BnqG064r5/aJrz2hOcECsQGyghoXwPrcp340Hxajle9RGrZcSXFoAnlhh/CbH+InxlB0n0+Eko3oGgtCrcgD+2FV7fQRL3bcQX15b4otqe7Hl0VLwgXiRQRiC2NYGE9qg33lnv8u342hwHEgfiAY/TVj0x4PFATGJwPPgnXvDEUOyJwVOcR6BdD3zeWHzB6eqNxVO2H03shMYmIt4Y1BOLiBcC5WhCe8RzaDni8QSX50XEGcdfBvFtgm08zmeD71V8VsXjTAPwxCLOjwLgzAdxfnPnR3SmE/whEWc+FW0q3q3SPjjOoddUmhfAyU3+S9UUzqSQC3Sv9DqV2p872+IFAkregVLyD5RSWFzOvpJytuwuRhAKispYv3MfCbFeioIb/X0Hyykq9VFS7qfEF6DsKPY+47weYr3OhmtfiY9ubRNIiPUQ6/UQ4/Wwbud+eiW3wusRWsfG4PUIndvEs7u4jB4dWxEb/HyMx8PuojJSOyYS6/EQ4xVivZ5D8/d68IoQE9yYVgzjY5w2McF5xHqdf/gikBDjDW7QObQhrtiIi5AQ68HrcZbjbvQ9te1LtECqUHYADu6BkkI4uBeKdsGWz8FfCiX7YO+3ENcGvlsN3ljYf4Q9q+JxNq7+MkjuB8X50HUIdOqOt7wYksc7y+/YB7xxzrI04Iy37uQMKzbe8UngiXXaBDegxLaGuFYQkwCxrSAmHvF43cV7qHny0ouJVOFMCvOBG0VkNk4CLGxp5xNUna6QzQVFbN1dzI7CEpZs3k2539mz/t93+4mP8VBc7j/UlVhNjEdo3yqWg2V++nVNol1iLN07JJKUEEN8jJf4WA+tYmNIjHM2ru1bxZEY5yUuxkMgoHRoHRfcMDsb3vgYD0nxscTHeoiP8bh7pSZEfKVQegAO7nY24v5y2POts7Es+AYCfti33XldUui027XW2aj6y+qeb/uewc/sg27p4PdBn9EQ3xaSukLrzs7GuVUyxMQHN+Zxzq5pYgdn41zx521xJc5MGIXsX4uIzAJGA51EJBeYAcQCqOojOA8eHw/kAMXANaGK5WipKtsLS1izfR9Lvt3N+p37yd1zkK27i2v0ISclxBDn9TA0tR2DuiVR4vPTr3MbOraOIy7GS6/kVrRvFUfreC8dWsfRJi7m2Nn7PVb4y6G4AAo2QMleZ6O+82tnmscLef9zNuo7VzZufnFtILEjtD0eOvSGzgMhJtEZJrRzugc69HY28O1SIbF9aNfPmHqE8uqjyxp4X4Gfhmr5R2N3URkrtu5hzfZ9rNi6l2Vb9rK7yNmri/UKA7om0bdza0YP6Ey3dgmktE/kuPaJ9OncmrYJsWGO3tTg9wU38t9A8W7I/dLZay8ucPbqYxMP7eUf3F3/vFp3hrYpzhFC9g+d/ujkfs6GPLGD85fQ3tnYJ7Z39uKNaUHsuLKSN1fu4OlPN7H02z3uidReya04a1AXhhzfll7JrTm5T0daxdnXFhHKDzp94cUFsG+b09VSuBV2fAV5653ul+J8OPBd7Z+PSXA28GX7ISXb6eNP7ud0w3ToGdxr7+D0tVd00xhzjLOtG7CzsITfvb6at1btpFObOH4yui8n904ms3t72iXann/YFe+G3CWwfZmTBDb8F/L/V3f7+LbO3n+rZOg1ytm4t01x9tw79oWOvaFVx+aL35gWJOqTwvIte7jhhWXsKCzhhtF9uXlMfxJi7dqIsCg/6Gzs92x2+vB3fg3frYHCLVXbdRsKfcdA+x7QLcPZwCe0d5JB5wHOFTLGmCMS1Unh9a+2c9Os5XgEnrrmRM4c2CXcIUWP4t2w+SPYu9Xp49+/E7Yurtom6XinGyf9Augx0unG6TSgynXbxpimFbVJYXN+ETfNWg7AGzeNYvDxbcMc0TGs9ABsfA+2LHZO6Oatr9r9E9vK6cc/6UfQZRB0GQJdB9sevzFhEJVJodwfYMb81QDMum6EJYSmVpQP3/wHtnwGW7+EvLWH3uvQG7oMhsET4bhMOH6Y0+dvjIkIUZkUpr/8NR/8L487vzeYkX2Twx1OyxbwQ9462PiBcyL420+dK4EqtE2FzMudJNBzpHOppjEmYkVdUij3B3h5WS79u7Rh2mm9wx1Oy7V7E6x+FT5+EEr3HZreZzQMOBcGToCepzjlD4wxLUbUJYXPNzo3J03MOj7MkbQwvjJY9RIsfwF2rTl0k5cnFk7/Pxg4DrplWkkFY1q4qPsfPHeJU637wuHWj92g8oPwv7edm8E+ftCZFpcEHXvBab+AXqc5l4R67V4OY44VUZcU1uzYR4dWsRzfPjHcoUSmsmJYvwBWzoVvFh6a3r4npF8EY+6yS0KNOYZFXVIoKvXRvaP1c1fh90HOIlj2LKx/89D07ifDSddDz1MhqZslA2OiQFQlhQOlPnYUlnDBsJRwhxIZDu5xTha/cQvu8416nw7DrnSuFoqJC2t4xpjmF1VJYcWWvQCkdIjirqMDebByNqx6GbY7N+/RobdT8TP7GrthzJgoF1VJYee+EgAGHxeFN6ttXw4LbnVKSoBTJ+j0W50jg56nOY9BNMZEvahKClsKigCi65xC7hJY9FunzhDAkAsg4xLnXgKPFf4zxlQVVUmhIPignOTWUdBXvm87PHfhoRITg74H5/4BOvQKa1jGmMgWVUlhb3G5++D4Y5IqbPoQlj4Nq19xpvUbCxf8C1pbOQ9jTMOiKims2l54bJ5POLALPpsJS550Sk6IF4ZfBYMmON1ExhjTSFGVFBJjvZSUB8IdRtNRhbdvgy8fg4DPmTbiBhj1KzsyMMYckahKCut27j82ah6VH4RlzzmlJ/Zvh67pcP7fnTLUx2rXmDGmWURVUoiL8eDza7jDODrLnoP5Nx56Pfo2pw6RPVTeGNMEoiYpqCplvgB9OrcOdyhHZtNH8PEDsOFd5/U59zglKCwZGGOaUNQkheIyPwAtrnMl4IdP/gb//Z3z+qTr4cTrnAfUG2NME4uapHCg1DkRGx/bQm7YUoWvX4K3bnVqFHUaCBf+yzlvYIwxIRI1SaHM51x11LVtQpgjaYTi3fCv06FwK7RKhrPuhNNusVIUxpiQi5qksLe4HHBONke00v3wz1Ng/w4Y9gMYfz/EtoBEZow5JkRNUijzO0cKgUAEX31UVgwzT3YSwsSZMGxquCMyxkSZqEkKFc8L6BCpdY/KS+De45zxrCssIRhjwiLC+1KajgYPECLy6qMDefCPk53x7ifDpH+ENx5jTNSKmiOFik6jiLvhtygf/jXK6TI65x4YeWPDnzHGmBAJ6ZGCiIwTkfUikiMi02t5v4eIvCciy0VkpYiMD1Ush44UIigr+MoOnUOY8Bc45aYIzFrGmGgSsqQgIl5gJnAeMBi4TEQGV2t2BzBXVYcBU4CQ9ZtoMCtEzDbX74NHR0NxvnN0cOK14Y7IGGNCeqRwEpCjqhtVtQyYDUys1kaBilrW7YDtoQrG7T4K1QIO14Jfwq7Vzh3KY+8OdzTGGAOE9pxCCrC10utc4ORqbX4L/EdEbgJaA2fXNiMRuR64HqBHjx5HF1UkZIVHz4Tty5znJI+7z25KM8ZEjFBujWrb/Fa/SeAy4GlVTQXGA8+JSI2YVPVRVc1W1ezOnTsfUTAaKbcnvP9HJyG0SoZfrLbnJBtjIkook0Iu0L3S61Rqdg9NA+YCqOpnQALQKRTBaDAfhfVE846v4P17wRMDP18FCcfgU+CMMS1aKJPCl0B/EektInE4J5LnV2uzBRgDICJpOEkhLyTRVFx9FK6c4C936hkBXPESxLUKUyDGGFO3kCUFVfUBNwILgbU4VxmtFpHfi8j5wWa/BK4Tka+AWcDVqqHp6AnrieaifPjnqc74KTdD3zPDEYUxxjQopDevqeoCYEG1aXdVGl8DnBrKGA4tyxlKcx8q7NsOfx0KgXLodzaM/X3zLt8YYw5D1F320qw5YdtSeGCwkxCyfwhTX46gGyWMMaamKCpz0cyXH/1vIfz7EvDEwvkPQdblzbt8Y4w5AtGTFJqzIN6ql+GlH4I3Hi6fDX3Pao6lGmPMUYuepBAchrz35qO/wH+D5w1u+AyS+4Z4gcYY03SiJyloM1x/VLjNSQjihVtzoFXH0C3LGGNCIGpONIf8SMFXBg9nO+Pj/2wJwRjTIkVNUiDU5xQ++SuUFzvF7U6cFqqlGGNMSEVPUggKyX0KgQC89wdn/JSbmn7+xhjTTKImKYT0ktS/D3eGw6+y+xCMMS1a9CSFUHUfff4o7NkE7bo79yMYY0wLFn1JoSmzQiAAb//aqXp67X+bcMbGGBMe0ZMUgsMmLZ39+BjQAIz6JSR1bbr5GmNMmERPUmjqZzT7ypyH5QCc+vMmmqkxxoRX9CSFppxZwA9PnuOMT/iLPRvBGHPMiJqkUHCgDGiiI4VFM2D7cujYB7LtngRjzLEjapJCQqyzql5PE2SFpc86wxs+t0tQjTHHlKhJChUSY71HN4M9m6G0EI4fDjFxTRKTMcZEiqhLCkdtyVPOcMQN4Y3DGGNCwJLC4cj/xqlx1CoZ0i8MdzTGGNPkLCkcjtd/5gwvfgY8R9kNZYwxEShqkoIe7TWp/nL49hNIPQl6j2qSmIwxJtJETVKocMR3NOf/zxmmfb/pgjHGmAgTdUnhiK2c4wy7pIU3DmOMCSFLCo216hVn2HdMeOMwxpgQsqTQGLs3QeFWiG8HHvvKjDHHLtvCNcYTY53hhf8KbxzGGBNiUZMUjvjio/07oSgPBo6Hgec1ZUjGGBNxoiYpVDjsUkWfBJ+mZgnBGBMFoi4pHBZfGSx92hkfOiWsoRhjTHOwpFCf/PVQXgRj77bid8aYqBDSpCAi40RkvYjkiMj0OtpcIiJrRGS1iPw7lPEctop7E3qMDG8cxhjTTGJCNWMR8QIzgbFALvCliMxX1TWV2vQHbgNOVdU9ItIlVPHokdS5+OwfzjDlhKYNxhhjIlQojxROAnJUdaOqlgGzgYnV2lwHzFTVPQCquiuE8Ryegg2gfuh7lt2bYIyJGqHc2qUAWyu9zg1Oq2wAMEBEPhGRxSIyrrYZicj1IrJERJbk5eWFKNxq3r3bGWZd0TzLM8aYCBDKpFDbxZ/V+3BigP7AaOAy4HERaV/jQ6qPqmq2qmZ37ty5yQOtVcDvDAd9r3mWZ4wxESCUSSEX6F7pdSqwvZY281S1XFU3AetxkkT4rZ0PXQZDbEK4IzHGmGYTyqTwJdBfRHqLSBwwBZhfrc1rwJkAItIJpztpYwhjapyyYmfoLw9vHMYY08xClhRU1QfcCCwE1gJzVXW1iPxeRM4PNlsIFIjIGuA94FZVLQhJPIfTeOdKZ3jyj0IRijHGRKyQXZIKoKoLgAXVpt1VaVyBW4J/zaJRZS4q7mLumh7KUIwxJuLYtZa12fSRM+wxIrxxGGNMM7OkUF0gAPtyIfXEI6ieZ4wxLZslhep2fuUMu2WENw5jjAmD6EkKjT3TvP5tZzjkgpCFYowxkSp6kkKQNNQl9OnfIS4Jeo1qnoCMMSaCRF1SqFdZsVMqO6mbnU8wxkSlw04KIuIVkWOzINCKF5xh+kXhjcMYY8KkzqQgIm1F5DYReVhEzhHHTTh3HF/SfCE2o+3LneGJ14Y3DmOMCZP6bl57DtgDfAZcC9wKxAETVXVFM8TW/PLWO8M2zVR0zxhjIkx9SaGPqmYAiMjjQD7QQ1X3N0tkTUwbc/nRtiXQaWDogzHGmAhV3zkFtxqcqvqBTS01IVRW5+njiiez9Ti5uUIxxpiIU9+RQqaI7OPQdjSx0mtV1bYhj645BXzOsH2P8MZhjDFhVGdSUFVvcwYSdv8L3rSW2CG8cRhjTBjVmRREJAH4MdAPWAk8GSyHfWz66C/OcMiF4Y3DGGPCqL5zCs8A2cDXwHjgL80SUYhofeeZAwHnctRuQ6FVx2aLyRhjIk195xQGV7r66Angi+YJKbRqvVG5ZK8z7Hlqs8ZijDGRprFXHx273UYAq152ht1PDG8cxhgTZvUdKWQFrzYC54qjY/fqo4o7mfufG944jDEmzOpLCl+p6rBmiyScCnMhtjXEtwl3JMYYE1b1dR8d1rPuW7Q9m0GsYKwxxtR3pNBFRG6p601VfSAE8YRMvRlu77fQ1Z60Zowx9SUFL9CGeipDtERSfXXKip1hl7TmD8YYYyJMfUlhh6r+vtkiCZfvVjvDrkPCG4cxxkSA+jrSj6kjhDoV5TnDLoPDG4cxxkSA+pLCmGaLIpxWveQMO/YJbxzGGBMB6kwKqrq7OQMJtTrLXBTkOMNO/ZotFmOMiVRRdx1mjTIXO76CXqPCEosxxkSaqEsKVfiD1Tu62eWoxhgD0Z4UKk4yx7UObxzGGBMhojspFG51hsl2PsEYYyDak8Lujc4w0Z6hYIwxEOKkICLjRGS9iOSIyPR62k0WERWR7FDForUVutjxlTNM7huqxRpjTIsSsqQgIl5gJnAeMBi4TERq3CEmIknAzcDnoYqlyvIqv9j4vjNs36M5Fm2MMREvlEcKJwE5qrpRVcuA2cDEWtrdDfwJKAlhLLXzxkG77uCNbfZFG2NMJAplUkgBtlZ6nRuc5hKRYUB3VX2jvhmJyPUiskREluTl5TVdhAEfHJfZdPMzxpgWLpRJobbaSW7Hvoh4gAeBXzY0I1V9VFWzVTW7c+fOTRfhd6vAU19NQGOMiS6hTAq5QPdKr1OB7ZVeJwHpwPsishkYAcwP1cnmGmUuDu51hv7yGm2NMSZahTIpfAn0F5HeIhIHTAHmV7ypqoWq2klVe6lqL2AxcL6qLglhTIeOX/LWO8PuJ4Z0ccYY05KELCmoqg+4EVgIrAXmqupqEfm9iJwfquU22v/edoY9TglvHMYYE0FC2qGuqguABdWm3VVH29GhjKWWJTqDlBOad7HGGBPBoveOZn85xLYCr51oNsaYCtGdFKLk4XLGGNNYUZMUahS5WPeGXY5qjDHVRE1SqCAVRwdtukDpvvAGY4wxESbqkoKrvAQGTQh3FMYYE1GiMyn4yyFvrXOi2RhjjCs6k0LBBmfYvnv97YwxJspET1KoXOdif7DaRue08MRijDERKnqSQpBUvgrVjhSMMaaKqEsKAAT8ztAuSTXGmCqiNCn4nKHHG944jDEmwkRnUvhutTMUSwrGGFNZdCaFA985w+S+4Y3DGGMiTNQkhSplLiouSY1rHY5QjDEmYkVNUqgg4CQD6zoyxpgaoi4pALBzJXRLD3cUxhgTcaIzKcS3hZLCcEdhjDERJzqTQkEOdBsa7iiMMSbiRE1SqFzlAlUo2Ru2WIwxJlJFTVKoIP5S8B2E44eFOxRjjIk4UZcUvAXrnZGEduENxBhjIlDUJQVPcYEz0mlAeAMxxpgIFH1JYfc3zkg7q5BqjDHVRV1SwBvvDJO6hTcOY4yJQFGTFLTi8qNAuTP0xIYvGGOMiVBRkxQqePdsCo7YsxSMMaa6qEsKbvdRXFJ44zDGmAgUdUnBs3cTxLUBT9StujHGNCg6t4xlB8IdgTHGRKSoSQpulQtV6DI4nKEYY0zECmlSEJFxIrJeRHJEZHot798iImtEZKWI/FdEeoYyHgDv3k3gtSuPjDGmNiFLCiLiBWYC5wGDgctEpPou+nIgW1WHAi8BfwpVPBU8ezdb2WxjjKlDKI8UTgJyVHWjqpYBs4GJlRuo6nuqWhx8uRhIDWE8Do8X2qaEfDHGGNMShTIppABbK73ODU6ryzTgrdreEJHrRWSJiCzJy8s7yrAEjss8ynkYY8yxKZRJQWqZprVMQ0SmAtnAn2t7X1UfVdVsVc3u3Lnz0UWlfpCoOb9ujDGHJZS39eYClavOpQLbqzcSkbOB3wBnqGppqIJxH7IT8IPH7mY2xpjahHKX+Uugv4j0FpE4YAowv3IDERkG/As4X1V3hTCWIHUesuPxhn5RxhjTAoUsKaiqD7gRWAisBeaq6moR+b2InB9s9megDfCiiKwQkfl1zK5JtKPIGSkrCuVijDGmxQppP4qqLgAWVJt2V6Xxs0O5/OoSKHNGOg9szsUaY0yLEVVnXDvLXmfEF7JTF8YY06JFTVJQwFNx8VPHPmGNxRhjIlXUJAWAGPzOiJ1oNsaYWkVVUoh1k4JdkmqMMbWJqqRwvOQ7IxoIbyDGGBOhoioplBGsjpp0XHgDMcaYCBVV/ShegkcIYucUWoLy8nJyc3MpKSkJdyjGtBgJCQmkpqYSG3tkjwiImqSgqngqkoKdaG4RcnNzSUpKolevXojUVkrLGFOZqlJQUEBubi69e/c+onlEVfeRmxSsIF6LUFJSQnJysiUEYxpJREhOTj6qo+sfRhn3AAAbRUlEQVSo2jq2luAXZUmhxbCEYMzhOdr/M1G1dewkwSeuxSaGNxBjjIlQUZUUDmq8MxKfFN5ATItQUFBAVlYWWVlZdOvWjZSUFPd1WVlZo+ZxzTXXsH79+sNe9oQJExg1alSVaVOnTuW1115zX/t8Ptq3b+++XrduHeeddx79+/cnLS2NKVOmsGtX44sPqyo33HAD/fr1IzMzkxUrVtTa7t///jcZGRkMGTKE2267zZ2+efNmzjrrLIYOHcqZZ57J9u1OpfylS5cyYsQI0tPTGTp0KC+99FKVZU6fPp0BAwaQlpbGzJkzAVi0aBHt2rVzv+8//OEP7mceeOABhgwZwpAhQ/j73//uTl++fDkjRowgIyODiRMncuDAAQB27drF6NGjad26NT//+c+rrMtpp53GwIED3eUUFBQA8PDDD5ORkUFWVhajRo1i3bp1AOTk5JCYmOi2/+lPfwrA3r173WlZWVkkJyfzq1/9qt55vf322wwfPpyMjAxOOOEE3n//fTeu6dOnk5qaWuX3re87blKq2qL+TjjhBD0Sj324Qf94+7WqM9qqlpcc0TxM81qzZk24Q3DNmDFD//znP9eYHggE1O/3N+my8vPztXv37tq/f3/99ttv3elXXHGFvvrqq+7r8vJybdeunaqqFhcXa9++ffXNN99031+0aNFhfYfz5s3T733ve6qq+tFHH+kpp5xSo813332nPXr00Pz8fA0EAnr55Zfr+++/r6qqkyZN0ueff15VVRcuXKhXX321qqquW7dOc3JyVFV169at2rVrV923b5+qqj766KN6zTXXaCAQcOevqvrOO+/oxIkTayx/+fLlOnToUC0uLtaysjIdPXq0bty4UVVVs7Ky9OOPP1ZV1X/961/629/+VlVV9+/frx9//LH+/e9/15/97GdV5nfqqafq8uXLayynsLDQHX/55Zd1woQJqqr6zTffaGZmZgPfpOrQoUP1k08+qXdeS5cu1e3bt6uq6ooVKzQ1NdVt9+mnn+rWrVvd37dCXd9xdbX97sASbcQ2NmquPoJKtY/sktQW53evr2bN9n1NOs/Bx7dlxveHHPbncnJymDRpEqeddhqff/45b7zxBr/73e9YtmwZBw8e5NJLL+Wuu5xiwKeddhoPP/ww6enpdOrUiR//+Me89dZbtGrVinnz5tGlS5ca83/ppZeYNGkS7dq1Y86cOdx6660NxvTcc89x+umnM378eHfamDFjDmu95s2bx5VXXunGvXPnTvLy8qj8tMMNGzaQlpZGcnIyAGeffTYvv/wyZ5xxBmvWrHGXOWbMGC6++GKeeuopBg48VJU4NTWV5ORk8vPzSUpK4p///CevvPKK2w9e2/dR2dq1axk5ciSJiU4X8Omnn86rr77KLbfcwoYNGzj11FMBGDt2LBMnTmTGjBm0adOGU089lbVr1zb6u2jbtq07XlRUdFj99GvXrqWwsJCRI0fWO6/hw4e70zMyMjhw4ADl5eXExsYycuRIfD5fjXnX9R03pajqPjqUFKJqtU0IrFmzhmnTprF8+XJSUlK47777WLJkCV999RXvvPMOa9asqfGZwsJCzjjjDL766itGjhzJk08+Weu8Z82axWWXXcZll13GrFmzGhXPqlWrOOGEE45qnbZt20b37ocelpiamsq2bduqtOnfvz+rVq1iy5YtlJeXM2/ePLZudR7FnpmZycsvvwzAyy+/zL59+ygsLKzy+U8//RSAXr16AbBp0yaef/55srOzGT9+PBs2bHDbfvzxx2RmZjJ+/Hj3+8zIyOD9999n9+7dFBUV8dZbb7nLHzRoEG+++SYAL774oju9IT/4wQ/Iysri3nvvrTL9oYceom/fvvzmN7/hr3/9qzs9JyeHYcOGMXr0aHd9Kps1axZTpkypkkjqmleFuXPncvLJJzd4b0FjvuOjFVVHCl6puCTVrmhpaY5kjz6U+vbty4knnui+njVrFk888QQ+n4/t27ezZs0aBg8eXOUziYmJnHfeeQCccMIJfPTRRzXmu23bNrZs2cKIESMQEfx+P+vWrWPQoEG17q025dVZqjUfoV59/p06dWLmzJlMnjyZmJgYTj75ZHJzcwF48MEHufHGG3niiSc444wz6NatGzExhzYx27Zt4+qrr+aFF15w51tSUkJSUhJLlixh7ty5XHvttbz33nuceOKJbN68mTZt2vD6669z4YUXsm7dOtLT07nllls4++yzadOmDcOGDXOX8fTTT/Ozn/2Mu+66i4kTJzbq5q05c+aQkpLCvn37uOCCC+jVqxeXX345ADfffDM333wzzz77LPfeey9PPPEEqampbNmyhY4dO/LFF19w0UUXsXbtWtq0aePOc/bs2bz44otVllPbvCp8/fXX3HHHHbzzzjsNxtvQd9wUomqXWQig4rGkYI5a69at3fFvvvmGv/3tb7z77rusXLmScePG1XqdeFxcnDvu9Xpr7R6YM2cOBQUF9O7dm169erFlyxZmz54NQHJyMnv27HHb7t69m06dOgEwZMgQli5d2mDc06dPJysri+zs7BrvpaamVtm7zs3N5fjjj6/RbuLEiXzxxRd8+umn9O/fn/79+wOQkpLCq6++yrJly/jd735HbGys+z0VFhYyYcIE/vjHP1ZJpikpKVx00UUAXHTRRe7J7Xbt2rkb2u9///scOHCAvXud56Fcf/31LFu2jA8//JB27dq5yx88eDDvvPMOS5cuZfLkyfTr16/B7yMlJQVwunguu+wyvvjiixptLr/8cl555RXAuVu4Y8eOAJx00kn07NmTnJwct+3SpUuJiYkhMzOz1uVVnhfAli1buPDCC3n++ecbdbNZfd9xU4mqpNBHdiJWDM80sX379pGUlETbtm3ZsWMHCxcuPOJ5zZo1i0WLFrF582Y2b97MF1984XYhjR49mtmzZ1NeXg44e8Znnnkm4HSBfPDBB7z99tvuvBYsWFCjG+u+++5jxYoVLFmypMayzz//fJ599lnA6brp2rVrlfMJFSquaNq9ezePPPII06ZNAyA/P9892rj33nu59tprASgtLWXixIlMmzaNCy64oMq8Jk2axLvvvgvAe++9x6BBgwDYuXOn22bx4sXExMS4V+JULH/z5s3MmzePSy+9tMr0QCDAPffcw49//OPav+Sg8vJy8vPz3fE333yT9PR0wEn0FV5//XX3vEheXh5+v1NtOScnh40bN1bZmFd0/VVW17z27NnDhAkTuP/++xkxYkS9sVao6ztuUo05Gx1Jf0d69dGjH2zQ5XdmOVcfmRYhUq8+qn4FSiAQ0KlTp2paWppOmDBBJ06cqM8995yqHrq6pfKVQqqqs2bN0mnTplVZRk5OjqamprpX4lTIyMjQJUuWqKrqnXfeqRkZGZqZmamTJ0/W/Px8t93q1av1nHPO0X79+mlaWppedtllumvXrkavo9/v1x/96Efap08fzcjI0GXLlqmqqs/n08r/7yZPnqxpaWmalpamc+bMcafPnj1b+/Xrp/3799frrrtOS0tLVVX1qaee0tjYWM3MzHT/Vq5cqaqqBQUFOm7cOE1PT9eRI0e60x988EEdPHiwZmZm6ogRI3Tx4sXuckaOHKlpaWmamZmp7733njv9/vvv1wEDBmj//v319ttvr/I9pqSkaIcOHbRNmzaakpKi69at03379unw4cM1IyND09LS9Oc//7l7JdkNN9zgLv+ss85y/y3OmTPHnT58+HB944033GUEAgHt3r27fvPNN1W+17rmNWPGDG3dunWV76Xi9/zFL36hKSkpKiKakpKid999d73fcXVHc/WRaC39iJEsOztba9vLachjH25k8DtTGdk2D8+t3zT8ARN2a9euJS0tLdxhGNPi1PZ/R0SWqmrNfsNqoqr7KEb8BJIHhDsMY4yJWFGVFLwErEKqMcbUI6qSwhDZbPcoGGNMPaJmC6koe2iD+OyBLcYYU5eoSQqocrzstnMKxhhTj6hJCrF+p2Iinqi6idsYYw5L1CSF1qV5AATa9whzJKalaIrS2QBPPvlklZuxqisrK6Njx47ceeedVaanpqa6d/GCU0560qRJ7us333yTE044gcGDBzNo0CB+/etfH8baOes3ZswY+vfvz7nnnltnDZ1f/epXpKenk56eXqXs9ZQpUxg4cCDp6elce+217h3ar7zyCkOHDiUrK4sTTzzRrQ9UXwntDRs2cNJJJ9GvXz8uv/xy9wa9P/3pT6SlpZGZmcnYsWOr3HFdV1xXXXUVmZmZZGRkcMkll1BUVATA448/TufOnd3fsKKQ3KZNmxg+fDhZWVmkp6fz2GOPufNqyjLhdcVVYfbs2YiIe1f3Z599RmZmJllZWWRmZjJ//vz6f9Cm0pibGSLp70hvXnvzlWdUZ7TVg0tnHdHnTfOL1JvXDldd5ZkrzJs3T0899VTt379/lekpKSm6Z88e93XlctIrVqzQfv366fr161XVKaP9j3/847Di+sUvfuGu091336233357jTavvfaannvuuerz+XT//v06fPhw3b9/v6qqvvnmm27p8MmTJ+ujjz6qqk6p6oobx5YuXapDhgxR1fpLaF9wwQX64osvqqrqtGnT3Hn997//1eLiYlVVfeihh/Tyyy9vMK7Kpapvuukmdx0fe+yxGqWzVVVLSkq0pKTE/Wz37t31u+++a/Iy4XXFVfHeqFGjNDs72/23UlRUpOXl5aqqum3bNu3SpUujy7Rb6exGceodaTs7UmiR3poOO79u2nl2y4Dz7juijz7zzDPMnDmTsrIyTjnlFB5++GECgQDXXHMNK1asQFW5/vrr6dq1KytWrODSSy8lMTGRL774okoNJHBKI9xyyy08+OCDfPnll1VqA9Xlj3/8I3feeScDBjjnyGJiYvjJT35yWOswb948Fi9eDDh7sePGjavyMBtwqsGOHj0ar9dLmzZtSE9P5z//+Q8XXnihW6ZbRDjppJPcwniVi8NVLhVdVwntVq1a8eGHH7pF5K666iruu+8+rrvuOs466yz3MyNGjHD3vOuLq6JUdSAQoKSkpMGigfHx8e54aWkpgUAAVWXjxo1NWia8vrhuv/12br/9du655x53WqtWrdzxgwcPArUXLWxqUdN9REXNI7sk1RylVatW8eqrr/Lpp5+yYsUKfD4fs2fPZunSpeTn5/P111+zatUqrrzySi699FKysrKYM2cOK1asqJEQioqK+OCDDxg/fnyzl8ouKChwaxulpKSwY8eOGm0yMzNZsGABBw8eJC8vjw8++KBGSeqysjJeeOEFxo0b50576aWXGDhwIJMmTeLxxx+vMd/KJbTz8vLo1KkTXq9zD1FtJbsBnnjiCbfKbENxXXnllXTr1o2NGzdyww03uNPnzp3L0KFDueSSS6osY/PmzQwdOpSePXtyxx130LVr1yYvE15XXF9++SW7du2q8v1VnseQIUPIzMzksccec7+jUIqaIwWPOkWs8EbNKh9bjnCPPhQWLVrEl19+6VYaPXjwIN27d+fcc89l/fr1/OxnP2P8+PGcc845Dc5r/vz5jB07loSEBC6++GKys7O5//778Xg8IS+V3Zh5jx8/niVLljBy5Ei6dOnCyJEja5Rq/vGPf8zZZ5/tPlQGYPLkyUyePJn33nuPO++8k//85z/ue9VLaNe291s9lmeeeYavv/6ahx56qFFxPfvss/j9fm644QZefPFFrrzySiZNmsQPfvAD4uPjmTlzJtdcc40bV69evVi5ciXbtm1j0qRJTJ48ucnLhNcW19SpU/nlL3/J888/X+tvcsopp7B69WpWr17ND3/4Q8aNG1djx6LJNaaP6Uj/gHHAeiAHmF7L+/HAnOD7nwO9GprnkZ5T+M/cf6rOaKvFW5Yd0edN84vUcwoPPPCA3nHHHbW2279/v86dO1fHjx+v1113narWf07h+9//vnbp0kV79uypPXv21ISEBLfIW2ZmpvuoSVWnGNu1116rqqpTpkzRZ555psG4x4wZo5mZmfqjH/2oxnt9+vRxC+Zt2bJFBw8e3OD8Lr74Yl24cKH7+o477tALL7yw3r7u1NRU99zI3r17NTMzU1955RX3fZ/Pp8nJyerz+VRV9cMPP9Tx48e777/11ls6ePBgzcvLa3RcFRYtWlTrYz3Lysq0ffv2tc5r6tSpVR57WmHmzJl622231ZhecR6iQm3rWFdcBQUFmpyc7P7+8fHxetxxx7nFCCs77bTT6j03VdnRnFMIWV+KiHiBmcB5wGDgMhEZXK3ZNGCPqvYDHgT+GKp42pQED409DT94w5j6nH322cydO9ctu1xQUMCWLVvIy8tDVbn44ovdx3MCJCUlsX///hrz2bNnD59//jm5ubluqeyHHnqoSqns5557DgCfz8cLL7zglsr+v//7P+655x63lr/f7+eBBx6osYxFixaxYsUKHnnkkRrvnX/++TzzzDOAsyc+ceLEGm18Ph+7d+8GYPny5axdu9btS3/kkUd4//33eeGFF/B4Dm1KcnJy3L3/iuKV7du3r7OEttfrZdSoUbz66qs1YlmyZAk//elPmT9/vvvsiPriCgQCbNy4EXB2eF9//XW3HHfl7rHXXnuNIUOcBzfl5ua6z78oKCjgs88+c8/VNFWZ8Lri6tixI/n5+e7vn52dzYIFCxg2bBibNm1yy3Rv2rSJnJwcevbsWeM3anKNyRxH8geMBBZWen0bcFu1NguBkcHxGCAfnMqtdf0d6ZHCJ8/9VnVGWy0qzG+4sYkIkXqkoKr6wgsvaGZmpmZkZOjw4cP1iy++0KVLl2pWVpZmZmZqVlaWu+c6Z84cHTBggGZmZlYpdfzYY4/pFVdcUWU5u3bt0s6dO2tZWZnu2bNHL730Uh06dKhmZGTor3/96yrloOfNm6fDhg3TQYMGaVpamk6fPv2w1mnXrl06evRo7devn44dO1Z3796tqqqLFy92jywOHDjglskeMWKEfvXVV6rq7N17vV7t27evW/b5nnvuUVXVP/zhD26p6JEjR7oPsK+vhPY333yj2dnZ2rdvX7300kvd7+mMM87Qrl27uu0nTZpUb1xlZWU6cuRITU9P1/T0dJ06dap79c+vfvUrHTJkiA4dOlTPPPNMXbdunao6RyLp6enu9/z444+731FTlQmvL67KKh9VPvXUU1XKdM+bN6/Rv21Els4WkcnAOFW9Nvj6B8DJqnpjpTargm1yg683BNvkV5vX9cD1AD169Djh22+/Pex4VrzzArpiFmk3vURCQsKRrpZpRlY625gjczSls0N51rW2M2LVM1Bj2qCqjwKPgvM8hSMJJmvsFTD2iiP5qDHGRI1QXp+ZC3Sv9DoV2F5XGxGJAdoBu0MYkzHGmHqEMil8CfQXkd4iEgdMAarfpz0fuCo4Phl4V0PVn2VaJPvnYMzhOdr/MyFLCqrqA27EOZm8FpirqqtF5Pcicn6w2RNAsojkALcA00MVj2l5EhISKCgosMRgTCOpKgUFBUd13jRqntFsWp7y8vIqlwsaYxqWkJBAamoqsbFVL7+PhBPNxhyV2NhYevfuHe4wjIkqVgjIGGOMy5KCMcYYlyUFY4wxrhZ3ollE8oDDv6XZ0QmnlEY0sXWODrbO0eFo1rmnqnZuqFGLSwpHQ0SWNObs+7HE1jk62DpHh+ZYZ+s+MsYY47KkYIwxxhVtSeHRcAcQBrbO0cHWOTqEfJ2j6pyCMcaY+kXbkYIxxph6WFIwxhjjOiaTgoiME5H1IpIjIjUqr4pIvIjMCb7/uYj0av4om1Yj1vkWEVkjIitF5L8i0gwPew2thta5UrvJIqIi0uIvX2zMOovIJcHferWI/Lu5Y2xqjfi33UNE3hOR5cF/3+PDEWdTEZEnRWRX8MmUtb0vIvJQ8PtYKSLDmzSAxjyzsyX9AV5gA9AHiAO+AgZXa3MD8EhwfAowJ9xxN8M6nwm0Co7/JBrWOdguCfgQWAxkhzvuZvid+wPLgQ7B113CHXczrPOjwE+C44OBzeGO+yjX+XRgOLCqjvfHA2/hPLlyBPB5Uy7/WDxSOAnIUdWNqloGzAYmVmszEXgmOP4SMEZEans0aEvR4Dqr6nuqWhx8uRjnSXgtWWN+Z4C7gT8Bx0L97cas83XATFXdA6Cqu5o5xqbWmHVWoG1wvB01n/DYoqjqh9T/BMqJwLPqWAy0F5Hjmmr5x2JSSAG2VnqdG5xWaxt1HgZUCCQ3S3Sh0Zh1rmwazp5GS9bgOovIMKC7qr7RnIGFUGN+5wHAABH5REQWi8i4ZosuNBqzzr8FpopILrAAuKl5Qgubw/3/fliOxecp1LbHX/2628a0aUkavT4iMhXIBs4IaUShV+86i4gHeBC4urkCagaN+Z1jcLqQRuMcDX4kIumqujfEsYVKY9b5MuBpVf2LiIwEnguucyD04YVFSLdfx+KRQi7QvdLrVGoeTrptRCQG55CzvsO1SNeYdUZEzgZ+A5yvqqXNFFuoNLTOSUA68L6IbMbpe53fwk82N/bf9jxVLVfVTcB6nCTRUjVmnacBcwFU9TMgAadw3LGqUf/fj9SxmBS+BPqLSG8RicM5kTy/Wpv5wFXB8cnAuxo8g9NCNbjOwa6Uf+EkhJbezwwNrLOqFqpqJ1Xtpaq9cM6jnK+qLflZro35t/0azkUFiEgnnO6kjc0aZdNqzDpvAcYAiEgaTlLIa9Yom9d84MrgVUgjgEJV3dFUMz/muo9U1SciNwILca5ceFJVV4vI74ElqjofeALnEDMH5whhSvgiPnqNXOc/A22AF4Pn1Leo6vlhC/ooNXKdjymNXOeFwDkisgbwA7eqakH4oj46jVznXwKPicgvcLpRrm7JO3kiMgun+69T8DzJDCAWQFUfwTlvMh7IAYqBa5p0+S34uzPGGNPEjsXuI2OMMUfIkoIxxhiXJQVjjDEuSwrGGGNclhSMMca4LCkY00gi4heRFZX+eonIaBEpDFboXCsiM4JtK09fJyL3hzt+YxrjmLtPwZgQOqiqWZUnBMuuf6Sq3xOR1sAKEamotVQxPRFYLiKvquonzRuyMYfHjhSMaSKqWgQsBfpWm34QWEETFi0zJlQsKRjTeImVuo5erf6miCTj1FhaXW16B5z6Qx82T5jGHDnrPjKm8Wp0HwWNEpHlQAC4L1iGYXRw+kpgYHD6zmaM1ZgjYknBmKP3kap+r67pIjIA+Dh4TmFFcwdnzOGw7iNjQkxV/wf8P+DX4Y7FmIZYUjCmeTwCnC4ivcMdiDH1sSqpxhhjXHakYIwxxmVJwRhjjMuSgjHGGJclBWOMMS5LCsYYY1yWFIwxxrgsKRhjjHH9f5lK1qTf4gePAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xeeJ8eVvLPpX"
   },
   "source": [
    "### [5.2.3] Feature Importance on TFIDF,<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Cj9XEmqQLPpY"
   },
   "source": [
    "#### [5.2.3.1] Top 10 important features of positive class from<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "oGRpCVJgLPpI"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['abdominal', 'ability', 'able', 'absence', 'absent', 'absolute', 'absolutely', 'absolutly', 'absorb']\n",
      "8083\n",
      "Top 10 Important features of positive class -\n",
      "        words      coef\n",
      "0      great  0.431203\n",
      "1       best  0.260459\n",
      "2  delicious  0.226270\n",
      "3       love  0.221095\n",
      "4       good  0.179635\n",
      "5    perfect  0.139171\n",
      "6      loves  0.137530\n",
      "7       find  0.105392\n",
      "8       nice  0.099771\n",
      "9   favorite  0.093241\n"
     ]
    }
   ],
   "source": [
    "lr = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = optimal_c)\n",
    "lr.fit(x_train,y_train) \n",
    "fn = tf_idf_vect.get_feature_names()\n",
    "print((fn[1:10]))\n",
    "w = list(lr.coef_[0])\n",
    "\n",
    "print(len(w))\n",
    "pos_coef =[]\n",
    "neg_coef = []\n",
    "pos_words =[]\n",
    "neg_words =[]\n",
    "\n",
    "for i,c in enumerate(w):\n",
    "#     print('i------------',i)\n",
    "#     print('c------------',c)\n",
    "    if c > 0:\n",
    "        pos_coef.append(c)\n",
    "        pos_words.append(fn[i]) \n",
    "    else:\n",
    "        neg_coef.append(abs(c))\n",
    "        neg_words.append(fn[i])\n",
    "\n",
    "pos_df = pd.DataFrame(columns= ['words', 'coef'])\n",
    "neg_df = pd.DataFrame(columns= ['words', 'coef'])\n",
    "pos_df['words'] = pos_words\n",
    "neg_df['words'] = neg_words\n",
    "pos_df['coef'] = pos_coef\n",
    "neg_df['coef'] = neg_coef\n",
    "\n",
    "pos_df = pos_df.sort_values('coef',axis = 0,ascending = False).reset_index(drop = True)\n",
    "\n",
    "print('Top 10 Important features of positive class -\\n',pos_df.head(10))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PowEfns4LPpg"
   },
   "source": [
    "#### [5.2.3.2] Top 10 important features of negative class from<font color='red'> SET 2</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "c9r7Zjx1LPpK",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 10 Important features of negative class -\n",
      "            words      coef\n",
      "0            not  0.260298\n",
      "1   disappointed  0.111554\n",
      "2          worst  0.099940\n",
      "3            bad  0.082465\n",
      "4       terrible  0.077616\n",
      "5          awful  0.075832\n",
      "6       horrible  0.075758\n",
      "7          money  0.072739\n",
      "8        thought  0.066702\n",
      "9  unfortunately  0.060907\n"
     ]
    }
   ],
   "source": [
    "neg_df = neg_df.sort_values('coef',axis = 0,ascending = False).reset_index(drop = True)\n",
    "\n",
    "print('Top 10 Important features of negative class -\\n',neg_df.head(10))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "sqYadGOgLPpi"
   },
   "source": [
    "## [5.3] Logistic Regression on AVG W2V,<font color='red'> SET 3</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "kByBCEA1LPpj"
   },
   "source": [
    "### [5.3.1] Applying Logistic Regression with L1 regularization on AVG W2V<font color='red'> SET 3</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Hyper parameter Tuning using Simple for loop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training w2v model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "no of words occured min 5 times  12483\n",
      "sample words  ['bought', 'apartment', 'infested', 'fruit', 'flies', 'hours', 'trap', 'attracted', 'many', 'within', 'days', 'practically', 'gone', 'may', 'not', 'long', 'term', 'solution', 'driving', 'crazy', 'consider', 'buying', 'one', 'caution', 'surface', 'sticky', 'try', 'avoid', 'touching', 'really', 'good', 'idea', 'final', 'product', 'outstanding', 'use', 'car', 'window', 'everybody', 'asks', 'made', 'two', 'thumbs', 'received', 'shipment', 'could', 'hardly', 'wait', 'love', 'call']\n"
     ]
    }
   ],
   "source": [
    "# w2v for train\n",
    "\n",
    "#Preparing Reviews for gensim model\n",
    "\n",
    "list_of_sentance_train = []\n",
    "for sentance in base_x_train :\n",
    "    list_of_sentance_train.append(sentance.split())\n",
    "    \n",
    "# Training w2v model\n",
    "w2v_model = Word2Vec(list_of_sentance_train , min_count = 5, size = 50, workers =4)\n",
    "w2v_words = list(w2v_model.wv.vocab)\n",
    "print('no of words occured min 5 times ',len(w2v_words))\n",
    "print(\"sample words \", w2v_words[0:50])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Converting Train data text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(43008, 50)\n",
      "[ 0.0695597   0.26210574 -0.03847536  0.67813659  0.21284877  0.28683027\n",
      "  0.03528085 -0.13975251  0.28017139  0.03001817 -0.35874062 -0.07485592\n",
      " -0.16230482  0.46536249  0.3186563  -0.17571915  0.14646051  0.45746915\n",
      " -0.50107873  0.59694558 -0.42665026  0.08409735 -0.34109022  0.20028363\n",
      "  0.02489534 -0.03462529 -0.51735375  0.23403109 -0.26359393  0.26219283\n",
      "  0.36722057  0.28272104 -0.20935628  0.2984979  -0.353033   -0.23942683\n",
      " -0.03639696 -0.15451757 -0.60611109  0.14132131  0.08259926  0.21612827\n",
      " -0.40216905 -0.12408005  0.41633012  0.40704982  0.43070303  0.11629507\n",
      "  0.43682186  0.08633048]\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Converting Reviews into Numerical Vectors using W2V vectors\n",
    "## Algorithm: Avg W2V\n",
    "\n",
    "# compute average word2vec for each review.\n",
    "sent_vectors = []; #the average word2vec for each sentance/review will store in this list\n",
    "# for sent in tqdm(list_of_sentance_train):\n",
    "for sent in (list_of_sentance_train):\n",
    "    sent_vec = np.zeros(50)\n",
    "    cnt_words = 0\n",
    "    for word in sent:\n",
    "        if word in w2v_words:\n",
    "            vec =  w2v_model.wv[word]\n",
    "            sent_vec += vec\n",
    "            cnt_words += 1\n",
    "    if cnt_words != 0 :\n",
    "        sent_vec /= cnt_words\n",
    "    sent_vectors.append(sent_vec)\n",
    "sent_vectors_train = np.array(sent_vectors)\n",
    "print(sent_vectors_train.shape)\n",
    "print(sent_vectors_train[0])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Converting CV data text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_of_sentance_cv = []\n",
    "for sentance in base_x_cv:\n",
    "    list_of_sentance_cv.append(sentance.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(18432, 50)\n",
      "[-6.12690673e-02  3.58439661e-01  4.19460560e-02  1.06436857e+00\n",
      " -8.50755068e-02  2.90837769e-01  1.43359740e-01 -8.00742883e-01\n",
      "  1.82480054e-01  1.10252766e-01  2.39786868e-01 -7.74416390e-03\n",
      "  3.20715215e-01  3.70796418e-01  6.15053054e-01 -3.46767233e-01\n",
      "  6.19271110e-01  3.28763154e-01 -9.80163909e-02  6.01097659e-01\n",
      "  5.01815096e-01  2.01231253e-01 -4.62018127e-01 -5.28082929e-04\n",
      "  1.63194770e-01 -9.94588276e-02 -7.13610587e-01  1.36658230e+00\n",
      " -8.12460679e-01  2.59664413e-02  6.92482793e-01  4.73151750e-01\n",
      " -1.33049624e-02 -6.72545285e-02 -3.82696081e-01  1.11811500e-02\n",
      "  1.48109242e-01  3.67342281e-01  1.93507934e-02  6.77744817e-01\n",
      "  2.22311611e-01  1.93007939e-01 -3.26650926e-01 -3.19799387e-01\n",
      "  4.11689929e-01  9.04277051e-01  7.30048991e-01  7.54266588e-01\n",
      "  1.85454310e-01  5.72548671e-01]\n"
     ]
    }
   ],
   "source": [
    "# average Word2Vec\n",
    "# compute average word2vec for each review.\n",
    "sent_vectors_cv = []; #the avg-w2v for each sentence/review is stored in this list\n",
    "# for sent in tqdm(list_of_sentance_cv):\n",
    "for sent in (list_of_sentance_cv):\n",
    "    \n",
    "    sent_vec = np.zeros(50)\n",
    "    cnt_words = 0 \n",
    "    for word in sent: #for each word in a review/sentance\n",
    "        if word in w2v_words:\n",
    "            vec = w2v_model.wv[word]\n",
    "            sent_vec += vec\n",
    "            cnt_words += 1\n",
    "    if cnt_words != 0:\n",
    "        sent_vec /= cnt_words\n",
    "    sent_vectors_cv.append(sent_vec)\n",
    "sent_vectors_cv = np.array(sent_vectors_cv)\n",
    "print(sent_vectors_cv.shape)\n",
    "print(sent_vectors_cv[0])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_of_sentance_test = []\n",
    "for sentance in base_x_test:\n",
    "    list_of_sentance_test.append(sentance.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(26331, 50)\n",
      "[-0.05361983  0.36880882 -0.30382451  0.62871976  0.3642346   0.28036152\n",
      " -0.19573003 -0.19000585  0.36496428 -0.07103195 -0.55244673 -0.55848462\n",
      " -0.044952   -0.37856722 -0.05389363 -0.22664909  0.56048387 -0.54759742\n",
      " -0.3410366   1.20927062 -0.44833872  0.42739542  0.59842996  0.12787154\n",
      " -0.43191077  0.48266059 -0.55542726  0.20326464 -0.22385902  0.91305132\n",
      "  0.11341208 -0.06216123 -0.37716007  0.31175344  0.00681467 -0.03707754\n",
      " -0.67906578 -0.0571281  -0.6105633   0.00714729 -0.56677026  0.45751253\n",
      " -0.1922163  -0.53992651  0.30422274 -0.25737675  0.36423105  0.19714763\n",
      "  0.11119163  0.66526588]\n"
     ]
    }
   ],
   "source": [
    "# average Word2Vec\n",
    "# compute average word2vec for each review.\n",
    "sent_vectors_test = []; #the avg-w2v for each sentence/review is stored in this list\n",
    "# for sent in tqdm(list_of_sentance_test):\n",
    "for sent in (list_of_sentance_test):\n",
    "\n",
    "    sent_vec = np.zeros(50)\n",
    "    cnt_words = 0 \n",
    "    for word in sent: #for each word in a review/sentance\n",
    "        if word in w2v_words:\n",
    "            vec = w2v_model.wv[word]\n",
    "            sent_vec += vec\n",
    "            cnt_words += 1\n",
    "    if cnt_words != 0:\n",
    "        sent_vec /= cnt_words\n",
    "    sent_vectors_test.append(sent_vec)\n",
    "sent_vectors_test = np.array(sent_vectors_test)\n",
    "print(sent_vectors_test.shape)\n",
    "print(sent_vectors_test[0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_train = sent_vectors_train\n",
    "x_cv = sent_vectors_cv\n",
    "x_test = sent_vectors_test\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l1',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',1//optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "O_8Gu2SfLPpn"
   },
   "source": [
    "### [5.3.2] Applying Logistic Regression with L2 regularization on AVG W2V,<font color='red'> SET 3</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GBbb9lxlLPpB"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l2',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',1//optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "pdXTwVIPLPpp"
   },
   "source": [
    "## [5.4] Logistic Regression on TFIDF W2V,<font color='red'> SET 4</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WP9NjwNSLPpp"
   },
   "source": [
    "### [5.4.1] Applying Logistic Regression with L1 regularization on TFIDF W2V,<font color='red'> SET 4</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Hyper parameter Tuning using Simple for loop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training w2v model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# w2v for train\n",
    "# Preparing Reviews for gensim model\n",
    "\n",
    "list_of_sentance_train = []\n",
    "for sentance in base_x_train:\n",
    "    list_of_sentance_train.append(sentance.split())\n",
    "\n",
    "w2v_model = Word2Vec(list_of_sentance_train , min_count = 5 ,size = 50, workers = 4)\n",
    "w2v_words = list(w2v_model.wv.vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'TfidfVectorizer' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-4cec63307e0f>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mtf_idf_vect\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mTfidfVectorizer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mngram_range\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mmin_df\u001b[0m\u001b[1;33m=\u001b[0m \u001b[1;36m10\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mmax_features\u001b[0m\u001b[1;33m=\u001b[0m \u001b[1;36m500\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[0mtf_idf_matrix\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf_idf_vect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbase_x_train\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mtfidf_feat\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtf_idf_vect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_feature_names\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mdictionary\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mzip\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtf_idf_vect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_feature_names\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlist\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtf_idf_vect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0midf_\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'TfidfVectorizer' is not defined"
     ]
    }
   ],
   "source": [
    "tf_idf_vect = TfidfVectorizer(ngram_range=(1,2),min_df= 10,max_features= 500)\n",
    "\n",
    "tf_idf_matrix = tf_idf_vect.fit_transform(base_x_train)\n",
    "tfidf_feat = tf_idf_vect.get_feature_names()\n",
    "dictionary = dict(zip(tf_idf_vect.get_feature_names(),list(tf_idf_vect.idf_)))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Converting Train data text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "tfidf_sent_vectors_train = []\n",
    "row = 0\n",
    "\n",
    "# for sent in tqdm(list_of_sentance_train):\n",
    "for sent in (list_of_sentance_train):\n",
    "    \n",
    "    sent_vec = np.zeros(50)\n",
    "    weight_sum = 0\n",
    "    for word in sent:\n",
    "        if word in w2v_words and word in tfidf_feat:\n",
    "            vec = w2v_model.wv[word]\n",
    "            tf_idf = dictionary[word]*(sent.count(word)/len(sent))\n",
    "            sent_vec += (vec * tf_idf)\n",
    "            weight_sum = tf_idf\n",
    "    if weight_sum != 0:\n",
    "        sent_vec /= weight_sum\n",
    "    tfidf_sent_vectors_train.append(sent_vec)\n",
    "    row += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Converting CV data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_of_sentance_cv = []\n",
    "for sentance in base_x_cv:\n",
    "    list_of_sentance_cv.append(sentance.split())\n",
    "    \n",
    "tfidf_sent_vectors_cv = []\n",
    "row = 0\n",
    "\n",
    "# for sent in tqdm(list_of_sentance_cv):\n",
    "for sent in (list_of_sentance_cv):\n",
    "    \n",
    "    sent_vec = np.zeros(50)\n",
    "    weight_sum = 0\n",
    "    for word in sent:\n",
    "        if word in w2v_words and word in tfidf_feat:\n",
    "            vec = w2v_model.wv[word]\n",
    "            tf_idf = dictionary[word]*(sent.count(word)/len(sent))\n",
    "            sent_vec += (vec * tf_idf)\n",
    "            weight_sum = tf_idf\n",
    "    if weight_sum != 0:\n",
    "        sent_vec /= weight_sum\n",
    "    tfidf_sent_vectors_cv.append(sent_vec)\n",
    "    row += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Converting test data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_of_sentance_test = []\n",
    "for sentance in base_x_test:\n",
    "    list_of_sentance_test.append(sentance.split())\n",
    "    \n",
    "tfidf_sent_vectors_test = []\n",
    "row = 0\n",
    "\n",
    "# for sent in tqdm(list_of_sentance_test):\n",
    "for sent in (list_of_sentance_test):\n",
    "    \n",
    "    sent_vec = np.zeros(50)\n",
    "    weight_sum = 0\n",
    "    for word in sent:\n",
    "        if word in w2v_words and word in tfidf_feat:\n",
    "            vec = w2v_model.wv[word]\n",
    "            tf_idf = dictionary[word]*(sent.count(word)/len(sent))\n",
    "            sent_vec += (vec * tf_idf)\n",
    "            weight_sum = tf_idf\n",
    "    if weight_sum != 0:\n",
    "        sent_vec /= weight_sum\n",
    "    tfidf_sent_vectors_test.append(sent_vec)\n",
    "    row += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_train = tfidf_sent_vectors_train\n",
    "x_cv = tfidf_sent_vectors_cv\n",
    "x_test = tfidf_sent_vectors_test\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l1',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',1//optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l1',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ynUK2tYDLPps"
   },
   "source": [
    "### [5.4.2] Applying Logistic Regression with L2 regularization on TFIDF W2V,<font color='red'> SET 4</font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GBbb9lxlLPpB"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#BOW\n",
    "\n",
    "C = [10**-3, 10**-2, 10**-1,10**0,10**1, 10**2, 10**3] #C = 1/lambda\n",
    "auc_train = []\n",
    "auc_cv = []\n",
    "\n",
    "for i in C :\n",
    "    model = LogisticRegression(penalty='l2',class_weight = 'balanced',C = i)\n",
    "    model.fit(x_train,y_train)\n",
    "    \n",
    "    y_train_prob = model.predict_proba(x_train)[:,1]\n",
    "    y_cv_prob =  model.predict_proba(x_cv)[:,1]\n",
    "    \n",
    "    auc_train.append(roc_auc_score(y_train,y_train_prob))\n",
    "    auc_cv.append(roc_auc_score(y_cv, y_cv_prob))\n",
    "    \n",
    "\n",
    "optimal_c = C[auc_cv.index(max(auc_cv))]\n",
    "C = [np.log10(x) for x in C]\n",
    "\n",
    "fig = plt.figure()\n",
    "\n",
    "plt.plot(C, auc_train, label='AUC Train')\n",
    "plt.scatter(C, auc_train, label='AUC Train')\n",
    "plt.plot(C,auc_cv,label = 'AUC CV')\n",
    "plt.scatter(C,auc_cv,label = 'AUC CV')\n",
    "\n",
    "plt.title('AUC Vs Hyperparameter')\n",
    "plt.xlabel('C = 1/lambda)')\n",
    "plt.ylabel('AUC')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# print('optimal lambda (Max AUC) is : ',1//optimal_c)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Testing with Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = LogisticRegression(penalty= 'l2',class_weight = 'balanced',C = optimal_c)\n",
    "model.fit(x_train,y_train)\n",
    "\n",
    "train_prob = model.predict_proba(x_train)[:,1]\n",
    "test_prob = model.predict_proba(x_test)[:,1]\n",
    "\n",
    "train_fpr,train_tpr,thresholds1 = metrics.roc_curve(y_train,train_prob)\n",
    "test_fpr,test_tpr,thresholds2 = metrics.roc_curve(y_test,test_prob)\n",
    "\n",
    "plt.plot(train_fpr,train_tpr, label = \"Train AUC - \" + str(auc(train_fpr,train_tpr)))\n",
    "plt.plot(test_fpr,test_tpr, label = 'Test AUC - ' + str(auc(test_fpr,test_tpr)))\n",
    "plt.xlabel('FPR')\n",
    "plt.ylabel('TPR')\n",
    "plt.title('ROC Error plot')\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix using Heatmap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix for train data\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#confusion matrix using heatmap for train data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_train, model.predict(x_train))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()\n",
    "\n",
    "#confusion matrix using heatmap for test data\n",
    "print('Confusion Matrix for train data')\n",
    "cm = confusion_matrix(y_test, model.predict(x_test))\n",
    "class_labels = ['Positive', 'Negative']\n",
    "df = pd.DataFrame(cm, index = class_labels, columns= class_labels)\n",
    "sb.heatmap(df, annot= True , fmt = 'd')\n",
    "plt.title('Confusion Matrix')\n",
    "plt.xlabel('Predicted')\n",
    "plt.ylabel('Actual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "crZ4qBCyDYN7"
   },
   "source": [
    "# [6] Conclusions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9OacHxi1DYN8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------------+----------------+---------------------+------+\n",
      "| Vectorizer | Regularization | Hyperameter(lambda) | AUC  |\n",
      "+------------+----------------+---------------------+------+\n",
      "|    BOW     |       l1       |         0.01        | 0.93 |\n",
      "|    BOW     |       l2       |        0.001        | 0.9  |\n",
      "|   TFIDF    |       l1       |         0.01        | 0.94 |\n",
      "|   TFIDF    |       l2       |        0.001        | 0.94 |\n",
      "|  AVG W2v   |       l1       |          1          | 0.89 |\n",
      "|  AVG W2v   |       l2       |          1          | 0.89 |\n",
      "| TFIDF W2v  |       l1       |        0.001        | 0.81 |\n",
      "| TFIDF W2v  |       l2       |        0.001        | 0.81 |\n",
      "+------------+----------------+---------------------+------+\n"
     ]
    }
   ],
   "source": [
    "# compare all your models \n",
    "from prettytable import PrettyTable\n",
    "x = PrettyTable()\n",
    "x.field_names = [\"Vectorizer\",\"Regularization\", \"Hyperameter(lambda)\", \"AUC\"]\n",
    "x.add_row([\"BOW\",\"l1\",0.01,0.93])\n",
    "x.add_row([\"BOW\",\"l2\",0.001,0.90])\n",
    "x.add_row([\"TFIDF\",\"l1\",0.01,0.94])\n",
    "x.add_row([\"TFIDF\",\"l2\",0.001,0.94])\n",
    "x.add_row([\"AVG W2v\",\"l1\",1,0.89])\n",
    "x.add_row([\"AVG W2v\",\"l2\",1,0.89])\n",
    "x.add_row([\"TFIDF W2v\",\"l1\",0.001,0.81])\n",
    "x.add_row([\"TFIDF W2v\",\"l2\",0.001,0.81])\n",
    "print(x)"
   ]
  },
  {
   "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}