Twitter User Gender Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ML-MAJOR-MAY-ML052B10-Adithya Ganesan\n",
    "#### Name: Adithya Ganesan\n",
    "#### Email: adithyaganesh2000@gmail.com\n",
    "#### MAY Batch  2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Problem Statement: \n",
    "For the given dataset, perform EDA with visualization, formulate 2 questions on the given data and answer the same. Then proceed to to build an ensemble classifier using 3 ML algorithms and find out which algorithm best suits the dataset with respect to the accuracy of the algorithm."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Solution\n",
    "1. The dataset is to be analysed and preliminary data cleaning is to be done.\n",
    "2. Data exploration and feature engineering are done for fine tuning of dataset.\n",
    "3. ML modelling and accuracy checking to find the optimal algorithm for the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Importing the required packages\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from datetime import datetime\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from collections import Counter\n",
    "import re\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn import linear_model, metrics\n",
    "from sklearn import naive_bayes\n",
    "from sklearn import svm\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preliminary Data Cleaning:\n",
    "The data is analysed and the irrelevant data is cleaned."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>_unit_id</th>\n",
       "      <th>_golden</th>\n",
       "      <th>_unit_state</th>\n",
       "      <th>_trusted_judgments</th>\n",
       "      <th>_last_judgment_at</th>\n",
       "      <th>gender</th>\n",
       "      <th>gender:confidence</th>\n",
       "      <th>profile_yn</th>\n",
       "      <th>profile_yn:confidence</th>\n",
       "      <th>created</th>\n",
       "      <th>...</th>\n",
       "      <th>profileimage</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>sidebar_color</th>\n",
       "      <th>text</th>\n",
       "      <th>tweet_coord</th>\n",
       "      <th>tweet_count</th>\n",
       "      <th>tweet_created</th>\n",
       "      <th>tweet_id</th>\n",
       "      <th>tweet_location</th>\n",
       "      <th>user_timezone</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>815719226</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:24</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12/5/13 1:48</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/414342229...</td>\n",
       "      <td>0</td>\n",
       "      <td>FFFFFF</td>\n",
       "      <td>Robbie E Responds To Critics After Win Against...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>110964</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>main; @Kan1shk3</td>\n",
       "      <td>Chennai</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>815719227</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:30</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10/1/12 13:51</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/539604221...</td>\n",
       "      <td>0</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>ÛÏIt felt like they were my friends and I was...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7471</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Eastern Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>815719228</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:33</td>\n",
       "      <td>male</td>\n",
       "      <td>0.6625</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11/28/14 11:30</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/657330418...</td>\n",
       "      <td>1</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>i absolutely adore when louis starts the songs...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5617</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>clcncl</td>\n",
       "      <td>Belgrade</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>815719229</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:10</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6/11/09 22:39</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/259703936...</td>\n",
       "      <td>0</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>Hi @JordanSpieth - Looking at the url - do you...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1693</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>Palo Alto, CA</td>\n",
       "      <td>Pacific Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>815719230</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/27/15 1:15</td>\n",
       "      <td>female</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4/16/14 13:23</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/564094871...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Watching Neighbours on Sky+ catching up with t...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31462</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    _unit_id  _golden _unit_state  _trusted_judgments _last_judgment_at  \\\n",
       "0  815719226    False   finalized                   3    10/26/15 23:24   \n",
       "1  815719227    False   finalized                   3    10/26/15 23:30   \n",
       "2  815719228    False   finalized                   3    10/26/15 23:33   \n",
       "3  815719229    False   finalized                   3    10/26/15 23:10   \n",
       "4  815719230    False   finalized                   3     10/27/15 1:15   \n",
       "\n",
       "   gender  gender:confidence profile_yn  profile_yn:confidence  \\\n",
       "0    male             1.0000        yes                    1.0   \n",
       "1    male             1.0000        yes                    1.0   \n",
       "2    male             0.6625        yes                    1.0   \n",
       "3    male             1.0000        yes                    1.0   \n",
       "4  female             1.0000        yes                    1.0   \n",
       "\n",
       "          created  ...                                       profileimage  \\\n",
       "0    12/5/13 1:48  ...  https://pbs.twimg.com/profile_images/414342229...   \n",
       "1   10/1/12 13:51  ...  https://pbs.twimg.com/profile_images/539604221...   \n",
       "2  11/28/14 11:30  ...  https://pbs.twimg.com/profile_images/657330418...   \n",
       "3   6/11/09 22:39  ...  https://pbs.twimg.com/profile_images/259703936...   \n",
       "4   4/16/14 13:23  ...  https://pbs.twimg.com/profile_images/564094871...   \n",
       "\n",
       "   retweet_count sidebar_color  \\\n",
       "0              0        FFFFFF   \n",
       "1              0        C0DEED   \n",
       "2              1        C0DEED   \n",
       "3              0        C0DEED   \n",
       "4              0             0   \n",
       "\n",
       "                                                text tweet_coord tweet_count  \\\n",
       "0  Robbie E Responds To Critics After Win Against...         NaN      110964   \n",
       "1  ÛÏIt felt like they were my friends and I was...         NaN        7471   \n",
       "2  i absolutely adore when louis starts the songs...         NaN        5617   \n",
       "3  Hi @JordanSpieth - Looking at the url - do you...         NaN        1693   \n",
       "4  Watching Neighbours on Sky+ catching up with t...         NaN       31462   \n",
       "\n",
       "    tweet_created      tweet_id   tweet_location               user_timezone  \n",
       "0  10/26/15 12:40  6.587300e+17  main; @Kan1shk3                     Chennai  \n",
       "1  10/26/15 12:40  6.587300e+17              NaN  Eastern Time (US & Canada)  \n",
       "2  10/26/15 12:40  6.587300e+17           clcncl                    Belgrade  \n",
       "3  10/26/15 12:40  6.587300e+17    Palo Alto, CA  Pacific Time (US & Canada)  \n",
       "4  10/26/15 12:40  6.587300e+17              NaN                         NaN  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Reading dataset and creating dataframe\n",
    "\n",
    "twitter = pd.read_csv('C:/Users/Adithya/Downloads/Information.csv', encoding = 'ISO-8859-1')\n",
    "twitter.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20050, 26)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twitter.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>_unit_id</th>\n",
       "      <th>_golden</th>\n",
       "      <th>_unit_state</th>\n",
       "      <th>_trusted_judgments</th>\n",
       "      <th>_last_judgment_at</th>\n",
       "      <th>gender</th>\n",
       "      <th>gender:confidence</th>\n",
       "      <th>profile_yn</th>\n",
       "      <th>profile_yn:confidence</th>\n",
       "      <th>created</th>\n",
       "      <th>...</th>\n",
       "      <th>profileimage</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>sidebar_color</th>\n",
       "      <th>text</th>\n",
       "      <th>tweet_coord</th>\n",
       "      <th>tweet_count</th>\n",
       "      <th>tweet_created</th>\n",
       "      <th>tweet_id</th>\n",
       "      <th>tweet_location</th>\n",
       "      <th>user_timezone</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>815719226</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:24</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12/5/13 1:48</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/414342229...</td>\n",
       "      <td>0</td>\n",
       "      <td>FFFFFF</td>\n",
       "      <td>Robbie E Responds To Critics After Win Against...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>110964</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>main; @Kan1shk3</td>\n",
       "      <td>Chennai</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>815719227</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:30</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10/1/12 13:51</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/539604221...</td>\n",
       "      <td>0</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>ÛÏIt felt like they were my friends and I was...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7471</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Eastern Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>815719228</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:33</td>\n",
       "      <td>male</td>\n",
       "      <td>0.6625</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11/28/14 11:30</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/657330418...</td>\n",
       "      <td>1</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>i absolutely adore when louis starts the songs...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5617</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>clcncl</td>\n",
       "      <td>Belgrade</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>815719229</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/26/15 23:10</td>\n",
       "      <td>male</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6/11/09 22:39</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/259703936...</td>\n",
       "      <td>0</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>Hi @JordanSpieth - Looking at the url - do you...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1693</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>Palo Alto, CA</td>\n",
       "      <td>Pacific Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>815719230</td>\n",
       "      <td>False</td>\n",
       "      <td>finalized</td>\n",
       "      <td>3</td>\n",
       "      <td>10/27/15 1:15</td>\n",
       "      <td>female</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>yes</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4/16/14 13:23</td>\n",
       "      <td>...</td>\n",
       "      <td>https://pbs.twimg.com/profile_images/564094871...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Watching Neighbours on Sky+ catching up with t...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31462</td>\n",
       "      <td>10/26/15 12:40</td>\n",
       "      <td>6.587300e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    _unit_id  _golden _unit_state  _trusted_judgments _last_judgment_at  \\\n",
       "0  815719226    False   finalized                   3    10/26/15 23:24   \n",
       "1  815719227    False   finalized                   3    10/26/15 23:30   \n",
       "2  815719228    False   finalized                   3    10/26/15 23:33   \n",
       "3  815719229    False   finalized                   3    10/26/15 23:10   \n",
       "4  815719230    False   finalized                   3     10/27/15 1:15   \n",
       "\n",
       "   gender  gender:confidence profile_yn  profile_yn:confidence  \\\n",
       "0    male             1.0000        yes                    1.0   \n",
       "1    male             1.0000        yes                    1.0   \n",
       "2    male             0.6625        yes                    1.0   \n",
       "3    male             1.0000        yes                    1.0   \n",
       "4  female             1.0000        yes                    1.0   \n",
       "\n",
       "          created  ...                                       profileimage  \\\n",
       "0    12/5/13 1:48  ...  https://pbs.twimg.com/profile_images/414342229...   \n",
       "1   10/1/12 13:51  ...  https://pbs.twimg.com/profile_images/539604221...   \n",
       "2  11/28/14 11:30  ...  https://pbs.twimg.com/profile_images/657330418...   \n",
       "3   6/11/09 22:39  ...  https://pbs.twimg.com/profile_images/259703936...   \n",
       "4   4/16/14 13:23  ...  https://pbs.twimg.com/profile_images/564094871...   \n",
       "\n",
       "   retweet_count sidebar_color  \\\n",
       "0              0        FFFFFF   \n",
       "1              0        C0DEED   \n",
       "2              1        C0DEED   \n",
       "3              0        C0DEED   \n",
       "4              0             0   \n",
       "\n",
       "                                                text tweet_coord tweet_count  \\\n",
       "0  Robbie E Responds To Critics After Win Against...         NaN      110964   \n",
       "1  ÛÏIt felt like they were my friends and I was...         NaN        7471   \n",
       "2  i absolutely adore when louis starts the songs...         NaN        5617   \n",
       "3  Hi @JordanSpieth - Looking at the url - do you...         NaN        1693   \n",
       "4  Watching Neighbours on Sky+ catching up with t...         NaN       31462   \n",
       "\n",
       "    tweet_created      tweet_id   tweet_location               user_timezone  \n",
       "0  10/26/15 12:40  6.587300e+17  main; @Kan1shk3                     Chennai  \n",
       "1  10/26/15 12:40  6.587300e+17              NaN  Eastern Time (US & Canada)  \n",
       "2  10/26/15 12:40  6.587300e+17           clcncl                    Belgrade  \n",
       "3  10/26/15 12:40  6.587300e+17    Palo Alto, CA  Pacific Time (US & Canada)  \n",
       "4  10/26/15 12:40  6.587300e+17              NaN                         NaN  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Removing all rows which don't have gender as 'male' or 'female' \n",
    "\n",
    "twitter = twitter.loc[twitter['gender'].isin(['female', 'male'])]\n",
    "twitter.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12894, 26)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twitter.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['_unit_id', '_golden', '_unit_state', '_trusted_judgments',\n",
       "       '_last_judgment_at', 'gender', 'gender:confidence', 'profile_yn',\n",
       "       'profile_yn:confidence', 'created', 'description', 'fav_number',\n",
       "       'gender_gold', 'link_color', 'name', 'profile_yn_gold', 'profileimage',\n",
       "       'retweet_count', 'sidebar_color', 'text', 'tweet_coord', 'tweet_count',\n",
       "       'tweet_created', 'tweet_id', 'tweet_location', 'user_timezone'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twitter.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above table, it is finalized that only the following columns are required for modelling the ML algorithm:\n",
    "1. 'gender'\n",
    "2. 'link_color'\n",
    "3. 'sidebar_color'\n",
    "4. 'text'\n",
    "5. 'description'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Dropping irrelevant columns from dataset\n",
    "\n",
    "twitter = twitter.drop(['_unit_id', '_golden', '_unit_state', '_last_judgment_at', 'gender:confidence', 'profile_yn', 'profile_yn:confidence', \n",
    "                        'created', 'fav_number', 'gender_gold', 'name', 'profile_yn_gold', 'profileimage', 'retweet_count', \n",
    "                        'tweet_coord', 'tweet_count', 'tweet_created', 'tweet_id', 'tweet_location', 'user_timezone', \n",
    "                        '_trusted_judgments'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>gender</th>\n",
       "      <th>description</th>\n",
       "      <th>link_color</th>\n",
       "      <th>sidebar_color</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>male</td>\n",
       "      <td>i sing my own rhythm.</td>\n",
       "      <td>08C2C2</td>\n",
       "      <td>FFFFFF</td>\n",
       "      <td>Robbie E Responds To Critics After Win Against...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>male</td>\n",
       "      <td>I'm the author of novels filled with family dr...</td>\n",
       "      <td>0084B4</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>ÛÏIt felt like they were my friends and I was...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>male</td>\n",
       "      <td>louis whining and squealing and all</td>\n",
       "      <td>ABB8C2</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>i absolutely adore when louis starts the songs...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>male</td>\n",
       "      <td>Mobile guy.  49ers, Shazam, Google, Kleiner Pe...</td>\n",
       "      <td>0084B4</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>Hi @JordanSpieth - Looking at the url - do you...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>female</td>\n",
       "      <td>Ricky Wilson The Best FRONTMAN/Kaiser Chiefs T...</td>\n",
       "      <td>3B94D9</td>\n",
       "      <td>0</td>\n",
       "      <td>Watching Neighbours on Sky+ catching up with t...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender                                        description link_color  \\\n",
       "0    male                              i sing my own rhythm.     08C2C2   \n",
       "1    male  I'm the author of novels filled with family dr...     0084B4   \n",
       "2    male                louis whining and squealing and all     ABB8C2   \n",
       "3    male  Mobile guy.  49ers, Shazam, Google, Kleiner Pe...     0084B4   \n",
       "4  female  Ricky Wilson The Best FRONTMAN/Kaiser Chiefs T...     3B94D9   \n",
       "\n",
       "  sidebar_color                                               text  \n",
       "0        FFFFFF  Robbie E Responds To Critics After Win Against...  \n",
       "1        C0DEED  ÛÏIt felt like they were my friends and I was...  \n",
       "2        C0DEED  i absolutely adore when louis starts the songs...  \n",
       "3        C0DEED  Hi @JordanSpieth - Looking at the url - do you...  \n",
       "4             0  Watching Neighbours on Sky+ catching up with t...  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twitter.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Exploration & Feature Engineering\n",
    "Here we are going to explore the relationships of the independent and dependent variables and modify the features to present a better dataset for the ML models.\n",
    "\n",
    "#### Questions to be answered at the end of EDA:\n",
    "1. How significant are the colors used by the users in classifying gender?\n",
    "2. Comment on the text used by users."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data exploration based on color"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#This function is to visualize the most used colors by each gender in their profile\n",
    "\n",
    "def explore_color_feature(df, feature):\n",
    "    '''\n",
    "    Definition of function for data exploration for the colors.\n",
    "    The explore_color_feature function plots the most used colors by gender as bar graphs.\n",
    "    \n",
    "    Arguments:\n",
    "    feature : 'sidebar_color', 'link_color'\n",
    "    df: dataset to be visualized.\n",
    "    \n",
    "    Returns:\n",
    "    None\n",
    "    '''\n",
    "\n",
    "    #Creating a subset of the gender and color of the feature being visualized\n",
    "    Color_subset = df.loc[:,['gender', feature]]\n",
    "\n",
    "    #Splitting the dataset w.r.t. 'gender'\n",
    "    Color_subset_male = Color_subset[Color_subset['gender'] == 'male']\n",
    "    Color_subset_female = Color_subset[Color_subset['gender'] == 'female']\n",
    "\n",
    "    #List of all colors and number of users for each gender\n",
    "    colors_list_male = list(Color_subset_male[feature].values.flatten())\n",
    "    colors_list_female = list(Color_subset_female[feature].values.flatten())\n",
    "\n",
    "    #30 Most common colors used by each gender (30 chosen for easy clarity in representation)\n",
    "    nCommon = 30\n",
    "    top_colors_female = [x[0] for x in Counter(colors_list_female).most_common(nCommon)]\n",
    "    top_colors_male = [x[0] for x in Counter(colors_list_male).most_common(nCommon)]\n",
    "\n",
    "    #Number of profiles using the each color in the 30 most common colors\n",
    "    top_colors_users_female = [x[1] for x in Counter(colors_list_female).most_common(nCommon)]\n",
    "    top_colors_users_male = [x[1] for x in Counter(colors_list_male).most_common(nCommon)]\n",
    "\n",
    "    #Deriving the hexadecimal value to represent each color in the graph\n",
    "    top_colors_female_hex = ['#' + x + '000000' for x in top_colors_female]\n",
    "    top_colors_female_hex = [x[0:7] for x in top_colors_female_hex]\n",
    "    top_colors_male_hex = ['#' + x + '000000' for x in top_colors_male]\n",
    "    top_colors_male_hex = [x[0:7] for x in top_colors_male_hex]\n",
    "\n",
    "    #Range of colors for each gender\n",
    "    color_range_female = list(range(len(top_colors_female)))\n",
    "    color_range_male = list(range(len(top_colors_male)))\n",
    "\n",
    "    #Color usage plot for female profiles\n",
    "    fig1, ax1 = plt.subplots()\n",
    "\n",
    "    bar_width = 0.5\n",
    "    rects1 = plt.barh(color_range_female, top_colors_users_female, bar_width, label = 'Female', color = top_colors_female_hex)\n",
    "    plt.yticks(color_range_female, top_colors_female_hex)\n",
    "    plt.xlabel('Color')\n",
    "    plt.ylabel(feature)\n",
    "    plt.title('Most used colors by Females for ' + feature)\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "    #Color usage plot for male profiles\n",
    "    fig2, ax2 = plt.subplots()\n",
    "\n",
    "    bar_width = 0.5\n",
    "    rects1 = plt.barh(color_range_male, top_colors_users_male, bar_width, label = 'Male', color = top_colors_male_hex)\n",
    "    plt.yticks(color_range_male, top_colors_male_hex)\n",
    "    plt.xlabel('Color')\n",
    "    plt.ylabel(feature)\n",
    "    plt.title('Most used colors by Males for ' + feature)\n",
    "    plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "#Data exploration for color features\n",
    "\n",
    "#Removing weird values from color\n",
    "twitter = twitter.loc[(twitter['sidebar_color']).str.contains('E\\+') != True]\n",
    "twitter = twitter.loc[(twitter['link_color']).str.contains('E\\+') != True]\n",
    "\n",
    "#Plot for the most used 'sidebar_colors'\n",
    "explore_color_feature(twitter, 'sidebar_color')\n",
    "\n",
    "#Plot for the most used 'link_colors'\n",
    "explore_color_feature(twitter, 'link_color')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Answer to question 1:\n",
    "As seen from plots displayed above, most users have not changed the default color of their profile, but if these are discarded, then there is sizeable dataset to be used for classification."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data exploration based on text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Function to normalize text\n",
    "\n",
    "def normalize_text_feature(txt):\n",
    "    '''\n",
    "    To normalize the text we convert them to strings, in lowercase and remove all punctuations.\n",
    "    \n",
    "    Arguments:\n",
    "    txt: the text in the dataset that needs to be normalized.\n",
    "    \n",
    "    Returns:\n",
    "    s: normalized string\n",
    "    '''\n",
    "    s = str(txt) \n",
    "    s = s.lower()\n",
    "    txt = re.sub('\\W\\s',' ',s)\n",
    "    s = re.sub('\\s\\W',' ',s)\n",
    "    s = re.sub('@',' at_the_rate ',s)\n",
    "    s = re.sub('\\s+',' ',s)\n",
    "    return s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Function to find the different words and the number of occurences of each word\n",
    "\n",
    "def compute_bag_of_words(txt):\n",
    "    '''    \n",
    "    Arguments:\n",
    "    txt: The text in the dataset that needs to be vectorised.\n",
    "    \n",
    "    Returns:\n",
    "    vectors: number of occurences of different words in txt \n",
    "    vocabulary: different words in txt\n",
    "    '''\n",
    "    vectorizer = CountVectorizer()\n",
    "    vectors = vectorizer.fit_transform(txt)\n",
    "    vocabulary = vectorizer.get_feature_names()\n",
    "    return vectors, vocabulary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Function to visualize the most used words in a dataset\n",
    "\n",
    "def print_most_frequent(bow, vocab, gender):\n",
    "    '''\n",
    "    The 30 most used words are selected and displayed in a graph.\n",
    "    \n",
    "    Arguments:\n",
    "    bow: bag of words - number of occurences of different words in the text\n",
    "    vocab: different words in the text\n",
    "    gender: the gender for which the graph is to be displayed\n",
    "    \n",
    "    Returns:\n",
    "    None\n",
    "    '''\n",
    "    color_idx = ['female', 'male']\n",
    "    label_table = [\"Most used words by females for 'all_text'\", \"Most used words by males for 'all_text'\"]\n",
    "    \n",
    "    #Selecting the 30 most used words to display\n",
    "    n_most = 30\n",
    "    idx = np.argsort(bow.sum(axis=0))\n",
    "    idx_most_used = np.zeros(n_most)\n",
    "    occurence_number = np.zeros(n_most)\n",
    "    words_most_used = [\"\" for x in range(n_most)]\n",
    "\n",
    "    for i in range(0,n_most):\n",
    "        idx_most_used[i] = idx[0, -1-i]\n",
    "        words_most_used[i] = vocab[np.int64(idx_most_used[i])]\n",
    "        occurence_number[i] = bow.sum(axis=0)[0, idx[0, -1-i]]\n",
    "\n",
    "    #Plotting the graph\n",
    "    fig, ax = plt.subplots()\n",
    "    \n",
    "    word_number = np.arange(n_most)+1\n",
    "    rects1 = plt.barh(word_number, occurence_number, label = label_table[color_idx.index(gender)])\n",
    "    plt.yticks(word_number,words_most_used)\n",
    "    plt.ylabel('Most used words')\n",
    "    plt.xlabel('Number of occurences')\n",
    "    plt.title(label_table[color_idx.index(gender)])\n",
    "    plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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>gender</th>\n",
       "      <th>link_color</th>\n",
       "      <th>sidebar_color</th>\n",
       "      <th>all_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>male</td>\n",
       "      <td>08C2C2</td>\n",
       "      <td>FFFFFF</td>\n",
       "      <td>robbie e responds to critics after win against...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>male</td>\n",
       "      <td>0084B4</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>ûïit felt like they were my friends and i was...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>male</td>\n",
       "      <td>ABB8C2</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>i absolutely adore when louis starts the songs...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>male</td>\n",
       "      <td>0084B4</td>\n",
       "      <td>C0DEED</td>\n",
       "      <td>hi jordanspieth looking at the url do you use ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>female</td>\n",
       "      <td>3B94D9</td>\n",
       "      <td>0</td>\n",
       "      <td>watching neighbours on sky+ catching up with t...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender link_color sidebar_color  \\\n",
       "0    male     08C2C2        FFFFFF   \n",
       "1    male     0084B4        C0DEED   \n",
       "2    male     ABB8C2        C0DEED   \n",
       "3    male     0084B4        C0DEED   \n",
       "4  female     3B94D9             0   \n",
       "\n",
       "                                            all_text  \n",
       "0  robbie e responds to critics after win against...  \n",
       "1  ûïit felt like they were my friends and i was...  \n",
       "2  i absolutely adore when louis starts the songs...  \n",
       "3  hi jordanspieth looking at the url do you use ...  \n",
       "4  watching neighbours on sky+ catching up with t...  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Data exploration for text features\n",
    "#Normalizing text and finding the most used words by each gender\n",
    "\n",
    "#Adding a new column to the dataframe to store normalised version of 'text' and 'description'\n",
    "twitter['text_norm'] = [normalize_text_feature(txt) for txt in twitter['text']]\n",
    "twitter['description_norm'] = [normalize_text_feature(txt) for txt in twitter['description']]\n",
    "\n",
    "#Concatenating the nomalized 'text_norm' and 'description_norm' into a single column as text can be represented in a single column\n",
    "twitter['all_text'] =twitter['text_norm'].str.cat(twitter['description_norm'],sep=' ')\n",
    "twitter = twitter.drop(['text', 'description', 'text_norm', 'description_norm'], axis = 1)\n",
    "\n",
    "#Extract separate gender dataframes\n",
    "female_data = twitter[twitter['gender']=='female']\n",
    "male_data = twitter[twitter['gender']=='male']\n",
    "\n",
    "#Computing and displaying bag of words and occurences of words for each gender\n",
    "male_bow, male_voc = compute_bag_of_words(male_data['all_text'])\n",
    "print_most_frequent(male_bow, male_voc, 'male')\n",
    "\n",
    "female_bow, female_voc = compute_bag_of_words(female_data['all_text'])\n",
    "print_most_frequent(female_bow, female_voc, 'female')\n",
    "\n",
    "twitter.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Answer to question 2:\n",
    "The most used words by the users are general words like 'and', 'the' etc. But the frequency of the words will be helpful in classifying the gender of the users."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and testing of ML algorithms\n",
    "The following classifiers have been chosen for training on the dataset:\n",
    "1. SGD Classifier\n",
    "2. Logistic Regression\n",
    "3. SVM Classifier\n",
    "\n",
    "The ML algorithms are trained on each feature of the dataset and the algorithm with the maximum accuracy is the most optimal model for this dataset and the feature that gives maximum accuracy is the optimal feature for classification of this data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Function to train and testing model\n",
    "\n",
    "def train_and_predict(df, model, modelname, feature):\n",
    "    '''\n",
    "    Trains model and computes accuracy of the model.\n",
    "    \n",
    "    Arguments:\n",
    "    df: dataset\n",
    "    model: ML algorithm to be used\n",
    "    model_name: name of ML algorithm\n",
    "    feature: dataset column on which model is trained\n",
    "    \n",
    "    Returns:\n",
    "    model: ML algorithm trained\n",
    "    full_voc: vocabulary of dataset\n",
    "    acc: accuracy of model\n",
    "    '''\n",
    "    print('Testing', modelname, \" on\", feature, \":\" )\n",
    "    \n",
    "    #Tokenization of data\n",
    "    full_bow, full_voc = compute_bag_of_words(df[feature])\n",
    "    \n",
    "    #Creating training and testing set\n",
    "    X = full_bow\n",
    "    y = LabelEncoder().fit_transform(df['gender'])\n",
    "    X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.2, random_state=0)\n",
    "    \n",
    "    #Training and prediction\n",
    "    model.fit(X_train,y_train)\n",
    "    y_pred = model.predict(X_test)\n",
    "    \n",
    "    #Computing Mean-squared-error\n",
    "    mse = metrics.mean_squared_error(y_test,y_pred)\n",
    "    print('mse: {:.4f}'.format(mse))\n",
    "    \n",
    "    #Computing accuracy\n",
    "    acc = round(model.score(X_test,y_test) * 100, 3)\n",
    "    print('score: ', acc, ' %')\n",
    "    \n",
    "    return model, full_voc, acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SGD Classifier  on all_text :\n",
      "mse: 0.3585\n",
      "score:  64.152  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SGD Classifier  on link_color :\n",
      "mse: 0.4028\n",
      "score:  59.72  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SGD Classifier  on sidebar_color :\n",
      "mse: 0.4533\n",
      "score:  54.666  %\n",
      "--------------------------------------------------------------------------------\n",
      "================================================================================\n",
      "\n",
      "================================================================================\n",
      "--------------------------------------------------------------------------------\n",
      "Testing Logistic Regression  on all_text :\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Adithya\\Anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:940: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mse: 0.3336\n",
      "score:  66.641  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing Logistic Regression  on link_color :\n",
      "mse: 0.4028\n",
      "score:  59.72  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing Logistic Regression  on sidebar_color :\n",
      "mse: 0.4530\n",
      "score:  54.705  %\n",
      "--------------------------------------------------------------------------------\n",
      "================================================================================\n",
      "\n",
      "================================================================================\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SVM Classifier  on all_text :\n",
      "mse: 0.3219\n",
      "score:  67.807  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SVM Classifier  on link_color :\n",
      "mse: 0.4067\n",
      "score:  59.331  %\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Testing SVM Classifier  on sidebar_color :\n",
      "mse: 0.4530\n",
      "score:  54.705  %\n",
      "--------------------------------------------------------------------------------\n",
      "================================================================================\n",
      "\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": [
    "#Training the models and logging the accuracy\n",
    "\n",
    "#List of the models\n",
    "model_list = [linear_model.SGDClassifier(), linear_model.LogisticRegression(), svm.SVC()]\n",
    "model_names = ['SGD Classifier', 'Logistic Regression', 'SVM Classifier']\n",
    "\n",
    "#List of all features to be used in prediction\n",
    "feature_name = ['all_text', 'link_color', 'sidebar_color']\n",
    "\n",
    "#A list to store all model accuracies\n",
    "accuracy_list = np.zeros(len(model_list) * len(model_names))\n",
    "\n",
    "#Filtered dataframe\n",
    "twitter_filter = twitter[['gender', 'link_color', 'sidebar_color', 'all_text']]\n",
    "\n",
    "#Training and calculating accuracy\n",
    "k = 0\n",
    "for i in range(0, len(model_list)):\n",
    "    \n",
    "    print('=' * 80)\n",
    "    \n",
    "    for j in range(0, len(feature_name)):\n",
    "        \n",
    "        print('-' * 80)\n",
    "        \n",
    "        model_name = model_names[i]\n",
    "        model_list[i], voc_color, accuracy_list[k] = train_and_predict(twitter_filter, model_list[i], model_name, feature_name[j])\n",
    "        k += 1\n",
    "        \n",
    "        print('-' * 80)\n",
    "    \n",
    "    print('=' * 80)\n",
    "    print()\n",
    "\n",
    "#Creating a list to store model names with prediction feature\n",
    "model_names_plot = [' ' for x in range(len(model_list) * len(model_names))]\n",
    "k = 0\n",
    "\n",
    "for i in range(0, len(model_list)):\n",
    "    \n",
    "    for j in range(0, len(feature_name)):\n",
    "        \n",
    "        model_names_plot[k] = model_names[i] + '-' + feature_name[j]\n",
    "        k += 1\n",
    "\n",
    "#Plotting the accuracy\n",
    "fig, ax1 = plt.subplots()\n",
    "ax1.set_xlim([0, 100])\n",
    "bar_width = 0.5\n",
    "model_number = np.arange(len(accuracy_list))\n",
    "rects1 = plt.barh(model_number, accuracy_list, bar_width, label = 'Brand Predictors')\n",
    "plt.yticks(model_number, model_names_plot)\n",
    "plt.xlabel('Accuracy')\n",
    "plt.ylabel('Model')\n",
    "plt.title('Accuracy of the different Classifiers')\n",
    "plt.tight_layout()\n",
    "for index, value in enumerate(accuracy_list):\n",
    "    plt.text(value, index, str(value))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusion\n",
    "As seen from the results above, the SVM Classifier gives the maximum accuracy when trained on the text of users in twitter. Hence, The SVM Classifier trained on text is the most optimal model among the considered algorithms and the text of users is the optimal feature for classifying gender in the given dataset."
   ]
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}