Ensemble (Phone) - 2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e9d50064",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\User\\anaconda3\\lib\\site-packages\\scipy\\__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.23.3\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    }
   ],
   "source": [
    "#Preprocessing\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import string\n",
    "import re\n",
    "import contractions\n",
    "import spacy\n",
    "import nltk\n",
    "from textblob import TextBlob\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.corpus import wordnet\n",
    "from cleantext import clean\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "#Model Algorithms\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import VotingClassifier,StackingClassifier\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import confusion_matrix,classification_report,accuracy_score\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn import model_selection\n",
    "import contractions\n",
    "\n",
    "import time\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf4db763",
   "metadata": {},
   "source": [
    "### Read Training Dataset "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "1bbb20f0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Initial :\n",
      "Train\n",
      "------------------\n",
      "       textID                                               text  \\\n",
      "0  cb774db0d1                I`d have responded, if I were going   \n",
      "1  549e992a42      Sooo SAD I will miss you here in San Diego!!!   \n",
      "2  088c60f138                          my boss is bullying me...   \n",
      "3  9642c003ef                     what interview! leave me alone   \n",
      "4  358bd9e861   Sons of ****, why couldn`t they put them on t...   \n",
      "\n",
      "                         selected_text sentiment Time of Tweet Age of User  \\\n",
      "0  I`d have responded, if I were going   neutral       morning        0-20   \n",
      "1                             Sooo SAD  negative          noon       21-30   \n",
      "2                          bullying me  negative         night       31-45   \n",
      "3                       leave me alone  negative       morning       46-60   \n",
      "4                        Sons of ****,  negative          noon       60-70   \n",
      "\n",
      "       Country  Population -2020  Land Area (Km²)  Density (P/Km²)  \n",
      "0  Afghanistan          38928346         652860.0               60  \n",
      "1      Albania           2877797          27400.0              105  \n",
      "2      Algeria          43851044        2381740.0               18  \n",
      "3      Andorra             77265            470.0              164  \n",
      "4       Angola          32866272        1246700.0               26  \n",
      "\n",
      "Test\n",
      "------------------\n",
      "                                                text sentiment\n",
      "0                Anyone iPhone 14 or Pro?\\n\\n#iPhone   neutral\n",
      "1  I remember the first time i bought a smart Pho...  positive\n",
      "2  🧵 iPhone 14 Pro Max, my take:\\n\\nAfter using t...  positive\n",
      "3  I have started saving for the next #iPhone wha...   neutral\n",
      "4  Mf #iphone after this fucking update everythin...  negative\n"
     ]
    }
   ],
   "source": [
    "train=pd.read_csv(\"train.csv\",encoding=\"unicode_escape\")\n",
    "test=pd.read_csv(\"iPhoneTest.csv\")\n",
    "print(\"Initial :\")\n",
    "\n",
    "print('Train')\n",
    "print('------------------')\n",
    "print(train.head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "print(test.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ec4cf45",
   "metadata": {},
   "source": [
    "### Attribute Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fea17548",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Attribute Selection :\n",
      "\n",
      "Train\n",
      "------------------\n",
      "                                                text sentiment\n",
      "0                I`d have responded, if I were going   neutral\n",
      "1      Sooo SAD I will miss you here in San Diego!!!  negative\n",
      "2                          my boss is bullying me...  negative\n",
      "3                     what interview! leave me alone  negative\n",
      "4   Sons of ****, why couldn`t they put them on t...  negative\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Attribute Selection :\n",
      "                                                text sentiment\n",
      "0                Anyone iPhone 14 or Pro?\\n\\n#iPhone   neutral\n",
      "1  I remember the first time i bought a smart Pho...  positive\n",
      "2  🧵 iPhone 14 Pro Max, my take:\\n\\nAfter using t...  positive\n",
      "3  I have started saving for the next #iPhone wha...   neutral\n",
      "4  Mf #iphone after this fucking update everythin...  negative\n"
     ]
    }
   ],
   "source": [
    "print('Attribute Selection :')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train=train[['text','sentiment']]\n",
    "\n",
    "print(train.head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test=test[['text','sentiment']]\n",
    "print('Attribute Selection :')\n",
    "print(test.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f36c5de5",
   "metadata": {},
   "source": [
    "### Drop rows with null values "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e5d6dbd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "train.isnull().sum()\n",
    "train=train.dropna(axis=0)\n",
    "\n",
    "test.isnull().sum()\n",
    "test=test.dropna(axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cca0e90c",
   "metadata": {},
   "source": [
    "### Convert to lowercase "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0a3aca1d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lowercase\n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Lowercase\n",
      "0                  anyone iphone 14 or pro?\\n\\n#iphone\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2    🧵 iphone 14 pro max, my take:\\n\\nafter using t...\n",
      "3    i have started saving for the next #iphone wha...\n",
      "4    mf #iphone after this fucking update everythin...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Lowercase')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text']=train['text'].str.lower()\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text']=test['text'].str.lower()\n",
    "print('Lowercase')\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79f9d156",
   "metadata": {},
   "source": [
    "### Remove URL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cb4ab375",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove URL\n",
      "\n",
      "Train\n",
      "------------------\n",
      "Remove URL: \n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                  anyone iphone 14 or pro?\\n\\n#iphone\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2    🧵 iphone 14 pro max, my take:\\n\\nafter using t...\n",
      "3    i have started saving for the next #iphone wha...\n",
      "4    mf #iphone after this fucking update everythin...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove URL')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].replace('((www.[^\\s]+)|(https?://[^\\s]+))', '', regex = True)\n",
    "print(\"Remove URL: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].replace('((www.[^\\s]+)|(https?://[^\\s]+))', '', regex = True)\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "962faba2",
   "metadata": {},
   "source": [
    "### Remove Emoji "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ff07ca6c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove Emoji\n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                  anyone iphone 14 or pro?\\n\\n#iphone\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2     iphone 14 pro max, my take:\\n\\nafter using th...\n",
      "3    i have started saving for the next #iphone wha...\n",
      "4    mf #iphone after this fucking update everythin...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "def remove_emojis(data):\n",
    "    emoj = re.compile(\"[\"\n",
    "        u\"\\U0001F600-\\U0001F64F\"  # emoticons\n",
    "        u\"\\U0001F300-\\U0001F5FF\"  # symbols & pictographs\n",
    "        u\"\\U0001F680-\\U0001F6FF\"  # transport & map symbols\n",
    "        u\"\\U0001F1E0-\\U0001F1FF\"  # flags (iOS)\n",
    "        u\"\\U00002500-\\U00002BEF\"  # chinese char\n",
    "        u\"\\U00002702-\\U000027B0\"\n",
    "        u\"\\U00002702-\\U000027B0\"\n",
    "        u\"\\U000024C2-\\U0001F251\"\n",
    "        u\"\\U0001f926-\\U0001f937\"\n",
    "        u\"\\U00010000-\\U0010ffff\"\n",
    "        u\"\\u2640-\\u2642\" \n",
    "        u\"\\u2600-\\u2B55\"\n",
    "        u\"\\u200d\"\n",
    "        u\"\\u23cf\"\n",
    "        u\"\\u23e9\"\n",
    "        u\"\\u231a\"\n",
    "        u\"\\ufe0f\"  # dingbats\n",
    "        u\"\\u3030\"\n",
    "                      \"]+\", re.UNICODE)\n",
    "    return re.sub(emoj, '', data)\n",
    "\n",
    "print('Remove Emoji')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(lambda text: remove_emojis(text))\n",
    "print(train['text'].head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(lambda text: remove_emojis(text))\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5fa10b9",
   "metadata": {},
   "source": [
    "### Remove @user "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "92c4583b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove @ User\n",
      "\n",
      "Train\n",
      "------------------\n",
      "Remove @user: \n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Remove @user: \n",
      "0                  anyone iphone 14 or pro?\\n\\n#iphone\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2     iphone 14 pro max, my take:\\n\\nafter using th...\n",
      "3    i have started saving for the next #iphone wha...\n",
      "4    mf #iphone after this fucking update everythin...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove @ User')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].replace('(@[^\\s]+)', '', regex = True)\n",
    "print(\"Remove @user: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].replace('(@[^\\s]+)', '', regex = True)\n",
    "print(\"Remove @user: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a1902f2",
   "metadata": {},
   "source": [
    "### Remove # keyword "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d10c4302",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove #\n",
      "\n",
      "Train\n",
      "------------------\n",
      "Remove #keyword: \n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Remove #keyword: \n",
      "0                         anyone iphone 14 or pro?\\n\\n\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2     iphone 14 pro max, my take:\\n\\nafter using th...\n",
      "3     i have started saving for the next  whatever pro\n",
      "4    mf  after this fucking update everything has b...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove #')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].replace('(#[^\\s]+)', '', regex = True)\n",
    "print(\"Remove #keyword: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].replace('(#[^\\s]+)', '', regex = True)\n",
    "print(\"Remove #keyword: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65d6bb81",
   "metadata": {},
   "source": [
    "### Remove numerical characters "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d1a1f4f7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove numerical characters\n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4     sons of ****, why couldn`t they put them on t...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Remove numbers: \n",
      "0                           anyone iphone  or pro?\\n\\n\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2     iphone  pro max, my take:\\n\\nafter using the ...\n",
      "3     i have started saving for the next  whatever pro\n",
      "4    mf  after this fucking update everything has b...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove numerical characters')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].replace('[0-9]+', '', regex = True)\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].replace('[0-9]+', '', regex = True)\n",
    "print(\"Remove numbers: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f28ffdfb",
   "metadata": {},
   "source": [
    "### Expand contractions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "9c91f652",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Expand contractions\n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                  i`d have responded, if i were going\n",
      "1        sooo sad i will miss you here in san diego!!!\n",
      "2                            my boss is bullying me...\n",
      "3                       what interview! leave me alone\n",
      "4    sons of ****, why couldn`t they put them on th...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                                anyone iphone or pro?\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2    iphone pro max, my take: after using the new i...\n",
      "3      i have started saving for the next whatever pro\n",
      "4    mf after this fucking update everything has be...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "def expand_contractions(text):\n",
    "    return \" \".join([contractions.fix(word) for word in str(text).split()])\n",
    "    \n",
    "    \n",
    "print('Expand contractions')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(lambda text: expand_contractions(text))\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(lambda text: expand_contractions(text))\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3144ac82",
   "metadata": {},
   "source": [
    "### Remove punctuation "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "af5c38e4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove punctuation\n",
      "\n",
      "Train\n",
      "------------------\n",
      "Remove punctuation: \n",
      "0                    id have responded if i were going\n",
      "1           sooo sad i will miss you here in san diego\n",
      "2                               my boss is bullying me\n",
      "3                        what interview leave me alone\n",
      "4    sons of  why couldnt they put them on the rele...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Remove punctuation: \n",
      "0                                 anyone iphone or pro\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2    iphone pro max my take after using the new i h...\n",
      "3      i have started saving for the next whatever pro\n",
      "4    mf after this fucking update everything has be...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove punctuation')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].str.translate(str.maketrans('', '', string.punctuation))\n",
    "print(\"Remove punctuation: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].str.translate(str.maketrans('', '', string.punctuation))\n",
    "print(\"Remove punctuation: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca1b07e6",
   "metadata": {},
   "source": [
    "### Remove repeated keywords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0c43e663",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove repeated keywords :\n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                    id have responded if i were going\n",
      "1            soo sad i will miss you here in san diego\n",
      "2                               my boss is bullying me\n",
      "3                        what interview leave me alone\n",
      "4    sons of  why couldnt they put them on the rele...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                                 anyone iphone or pro\n",
      "1    i remember the first time i bought a smart pho...\n",
      "2    iphone pro max my take after using the new i h...\n",
      "3      i have started saving for the next whatever pro\n",
      "4    mf after this fucking update everything has be...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print('Remove repeated keywords :')\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].replace(r'(.)\\1+', r'\\1\\1', regex = True)\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].replace(r'(.)\\1+', r'\\1\\1', regex = True)\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0883cc7a",
   "metadata": {},
   "source": [
    "###  Remove Stopwords "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6bc7d6d9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove stopwords: \n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                          id responded going\n",
      "1                      soo sad miss san diego\n",
      "2                               boss bullying\n",
      "3                       interview leave alone\n",
      "4    sons couldnt put releases already bought\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                                    anyone iphone pro\n",
      "1    remember first time bought smart phone g showe...\n",
      "2    iphone pro max take using new noticed interest...\n",
      "3                     started saving next whatever pro\n",
      "4    mf fucking update everything become slow canno...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "stop_words = set(stopwords.words('english'))\n",
    "def cleaning_stopwords(text):\n",
    "    return \" \".join([word for word in str(text).split() if word not in stop_words])\n",
    "\n",
    "print(\"Remove stopwords: \")\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(lambda text: cleaning_stopwords(text))\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(lambda text: cleaning_stopwords(text))\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4714f948",
   "metadata": {},
   "source": [
    "### Tokenization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "efb095b8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tokenization: \n",
      "\n",
      "Train\n",
      "------------------\n",
      "Tokenization: \n",
      "0                             [id, responded, going]\n",
      "1                       [soo, sad, miss, san, diego]\n",
      "2                                   [boss, bullying]\n",
      "3                          [interview, leave, alone]\n",
      "4    [sons, couldnt, put, releases, already, bought]\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "Tokenization: \n",
      "0                                [anyone, iphone, pro]\n",
      "1    [remember, first, time, bought, smart, phone, ...\n",
      "2    [iphone, pro, max, take, using, new, noticed, ...\n",
      "3               [started, saving, next, whatever, pro]\n",
      "4    [mf, fucking, update, everything, become, slow...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(\"Tokenization: \")\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(nltk.word_tokenize)\n",
    "print(\"Tokenization: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(nltk.word_tokenize)\n",
    "print(\"Tokenization: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "de8d323f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Train\n",
      "------------------\n",
      "                                              text sentiment\n",
      "0                           [id, responded, going]   neutral\n",
      "1                     [soo, sad, miss, san, diego]  negative\n",
      "2                                 [boss, bullying]  negative\n",
      "3                        [interview, leave, alone]  negative\n",
      "4  [sons, couldnt, put, releases, already, bought]  negative\n",
      "\n",
      "Test\n",
      "------------------\n",
      "                                                text sentiment\n",
      "0                              [anyone, iphone, pro]   neutral\n",
      "1  [remember, first, time, bought, smart, phone, ...  positive\n",
      "2  [iphone, pro, max, take, using, new, noticed, ...  positive\n",
      "3             [started, saving, next, whatever, pro]   neutral\n",
      "4  [mf, fucking, update, everything, become, slow...  negative\n"
     ]
    }
   ],
   "source": [
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "print(train.head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "print(test.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1cc5236d",
   "metadata": {},
   "source": [
    "### Label Encoding "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "021e4e6b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Train\n",
      "------------------\n",
      "                                              text  sentiment\n",
      "0                           [id, responded, going]          0\n",
      "1                     [soo, sad, miss, san, diego]         -1\n",
      "2                                 [boss, bullying]         -1\n",
      "3                        [interview, leave, alone]         -1\n",
      "4  [sons, couldnt, put, releases, already, bought]         -1\n",
      "\n",
      "Test\n",
      "------------------\n",
      "                                                text  sentiment\n",
      "0                              [anyone, iphone, pro]          0\n",
      "1  [remember, first, time, bought, smart, phone, ...          1\n",
      "2  [iphone, pro, max, take, using, new, noticed, ...          1\n",
      "3             [started, saving, next, whatever, pro]          0\n",
      "4  [mf, fucking, update, everything, become, slow...         -1\n"
     ]
    }
   ],
   "source": [
    "train['sentiment'] = train['sentiment'].map({\"positive\": 1, \"neutral\": 0, \"negative\": -1})\n",
    "test['sentiment'] = test['sentiment'].map({\"positive\": 1, \"neutral\": 0, \"negative\": -1})\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "print(train.head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "print(test.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "045ff26a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Train\n",
      "------------------\n",
      " 0    11117\n",
      " 1     8582\n",
      "-1     7781\n",
      "Name: sentiment, dtype: int64\n",
      "\n",
      "Test\n",
      "------------------\n",
      "-1    126\n",
      " 1    107\n",
      " 0     77\n",
      "Name: sentiment, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "print(train['sentiment'].value_counts())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "print(test['sentiment'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1fe05071",
   "metadata": {},
   "source": [
    "### POS "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "55c328fb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "POS : \n",
      "\n",
      "Train\n",
      "------------------\n",
      "POS tagging: \n",
      "0           [(id, NN), (responded, VBD), (going, VBG)]\n",
      "1    [(soo, JJ), (sad, JJ), (miss, NN), (san, NN), ...\n",
      "2                        [(boss, IN), (bullying, VBG)]\n",
      "3         [(interview, NN), (leave, VBP), (alone, RB)]\n",
      "4    [(sons, NNS), (couldnt, VBP), (put, VBN), (rel...\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "POS tagging: \n",
      "0              [(anyone, NN), (iphone, CD), (pro, NN)]\n",
      "1    [(remember, VB), (first, JJ), (time, NN), (bou...\n",
      "2    [(iphone, JJ), (pro, JJ), (max, NNS), (take, V...\n",
      "3    [(started, VBN), (saving, VBG), (next, JJ), (w...\n",
      "4    [(mf, NN), (fucking, VBG), (update, JJ), (ever...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "print(\"POS : \")\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(nltk.pos_tag)\n",
    "print(\"POS tagging: \")\n",
    "print(train['text'].head())\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(nltk.pos_tag)\n",
    "print(\"POS tagging: \")\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ee368aa",
   "metadata": {},
   "source": [
    "### Lemmatization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e2ae3553",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lemmatization : \n",
      "\n",
      "Train\n",
      "------------------\n",
      "0                          id respond go\n",
      "1                 soo sad miss san diego\n",
      "2                              bos bully\n",
      "3                  interview leave alone\n",
      "4    son couldnt put release already buy\n",
      "Name: text, dtype: object\n",
      "\n",
      "Test\n",
      "------------------\n",
      "0                                    anyone iphone pro\n",
      "1    remember first time buy smart phone g show gra...\n",
      "2    iphone pro max take use new notice interesting...\n",
      "3                         start save next whatever pro\n",
      "4    mf fuck update everything become slow can not ...\n",
      "Name: text, dtype: object\n"
     ]
    }
   ],
   "source": [
    "def get_wordnet_pos(treebank_pos):\n",
    "    if treebank_pos.startswith('J'):\n",
    "        return wordnet.ADJ\n",
    "    elif treebank_pos.startswith('V'):\n",
    "        return wordnet.VERB\n",
    "    elif treebank_pos.startswith('N'):\n",
    "        return wordnet.NOUN\n",
    "    elif treebank_pos.startswith('R'):\n",
    "        return wordnet.ADV\n",
    "    else:\n",
    "        return wordnet.NOUN\n",
    "\n",
    "def lemmatization(text):\n",
    "    lemmatized_tokens = []\n",
    "\n",
    "    for i in range(len(text)):\n",
    "        word, pos = text[i]\n",
    "        \n",
    "        lemmatized_tokens.append(nltk.WordNetLemmatizer().lemmatize(word, get_wordnet_pos(pos)))\n",
    "        \n",
    "    return \" \".join(lemmatized_tokens)\n",
    "\n",
    "print(\"Lemmatization : \")\n",
    "print('\\nTrain')\n",
    "print('------------------')\n",
    "train['text'] = train['text'].apply(lambda text: lemmatization(text))\n",
    "print(train['text'].head())\n",
    "\n",
    "\n",
    "print('\\nTest')\n",
    "print('------------------')\n",
    "test['text'] = test['text'].apply(lambda text: lemmatization(text))\n",
    "print(test['text'].head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "daefa7a9",
   "metadata": {},
   "source": [
    "### Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "483b3029",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train=train['text']\n",
    "X_test=test['text']\n",
    "y_train=train['sentiment']\n",
    "y_test=test['sentiment']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "50c0db16",
   "metadata": {},
   "outputs": [],
   "source": [
    "tfidf_vectorizer= TfidfVectorizer(max_df=0.90, min_df=2,max_features =1000)\n",
    "X_train = tfidf_vectorizer.fit_transform(X_train) #Tokenize and build vocabulary\n",
    "X_test = tfidf_vectorizer.transform(X_test) #Tokenize and build vocabulary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0372f97e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(27480, 1000)\n",
      "(27480,)\n",
      "(310, 1000)\n",
      "(310,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(y_train.shape)\n",
    "print(X_test.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "357e5384",
   "metadata": {},
   "source": [
    "## Ensemble Model "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e4713429",
   "metadata": {},
   "source": [
    "#### Confusion Matrix "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "20f88a8e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def confusion_matrix_evaluation(classifier,y_predicted):\n",
    "    mat=confusion_matrix(y_test,y_predicted)\n",
    "    sns.heatmap(mat.T, xticklabels=['Predicted 1','Predicted 0','Predicted -1'],yticklabels=['Actual 1','Actual 0','Actual -1'],\n",
    "                annot=True, fmt='d', cbar=False,cmap=\"YlGnBu\")\n",
    "    plt.title('Confusion Matrix for '+classifier)\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    plt.savefig(\"confusion_matrix.png\")\n",
    "    plt.show()\n",
    "    print(classifier+' Classification Report ')\n",
    "    print('======================================================')\n",
    "    print(classification_report(y_test,y_predicted))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84cd7cd3",
   "metadata": {},
   "source": [
    "#### Define Base Learners "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "700bfedf",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "m1=SVC(C=0.1,kernel='linear',random_state=1)\n",
    "m2=MultinomialNB(alpha=20)\n",
    "m3=RandomForestClassifier(random_state=1,max_leaf_nodes=65,max_depth=100)\n",
    "m4=LogisticRegression(multi_class='multinomial',random_state=1,solver='saga')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7b231fe5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_accuracy(m1,m2,m3,m4,classifier):\n",
    "    print('Doing 2-fold Cross Validation Here :')\n",
    "    for iterclf,iterlabel in zip([m1,m2,m3,m4,classifier],\n",
    "                             ['SVC',\n",
    "                              'MNB',\n",
    "                              'Random Forest Tree',\n",
    "                              'Logistic Regression',\n",
    "                               classifier\n",
    "                             ]):\n",
    "     start_time = time.time()\n",
    "     scores=model_selection.cross_val_score(iterclf,X_train,y_train,cv=2,scoring='accuracy')\n",
    "     run_time = time.time() - start_time\n",
    "     print('Accuracy :%0.6f  (+/- %0.4f)  [%20s]   ||  %4.2f seconds'%(scores.mean(),scores.std(),iterlabel,run_time))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "41a23cf1",
   "metadata": {},
   "source": [
    "### Voting "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "1c8bfb71",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Doing 2-fold Cross Validation Here :\n",
      "Accuracy :0.627147  (+/- 0.0008)  [                 SVC]   ||  42.29 seconds\n",
      "Accuracy :0.592722  (+/- 0.0013)  [                 MNB]   ||  0.02 seconds\n",
      "Accuracy :0.665102  (+/- 0.0028)  [  Random Forest Tree]   ||  2.68 seconds\n",
      "Accuracy :0.683333  (+/- 0.0010)  [ Logistic Regression]   ||  0.20 seconds\n",
      "Accuracy :0.650255  (+/- 0.0021)  [VotingClassifier(estimators=[('SVC',\n",
      "                              SVC(C=0.1, kernel='linear', random_state=1)),\n",
      "                             ('Multinomial Naive Bayes',\n",
      "                              MultinomialNB(alpha=20)),\n",
      "                             ('Random Forest Tree',\n",
      "                              RandomForestClassifier(max_depth=100,\n",
      "                                                     max_leaf_nodes=65,\n",
      "                                                     random_state=1)),\n",
      "                             ('Logistic Regression',\n",
      "                              LogisticRegression(multi_class='multinomial',\n",
      "                                                 random_state=1,\n",
      "                                                 solver='saga'))])]   ||  40.32 seconds\n",
      "\t\t\t Voting\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ensemble Method - Voting Classification Report \n",
      "======================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          -1       0.79      0.15      0.25       126\n",
      "           0       0.29      0.96      0.45        77\n",
      "           1       0.85      0.27      0.41       107\n",
      "\n",
      "    accuracy                           0.39       310\n",
      "   macro avg       0.65      0.46      0.37       310\n",
      "weighted avg       0.69      0.39      0.36       310\n",
      "\n",
      "Training Accuracy : 0.6861    ||   Training Time  48.28 seconds\n"
     ]
    }
   ],
   "source": [
    "votingCLF=VotingClassifier([('SVC',m1),('Multinomial Naive Bayes',m2),('Random Forest Tree',m3),('Logistic Regression',m4)\n",
    "                       ])\n",
    "model_accuracy(m1,m2,m3,m4,votingCLF)  \n",
    "start_time = time.time()\n",
    "votingCLF.fit(X_train,y_train)\n",
    "run_time = time.time() - start_time\n",
    "predicted_Vote=votingCLF.predict(X_test)\n",
    "\n",
    "print('\\t\\t\\t Voting')\n",
    "print('================================================================\\n')\n",
    "confusion_matrix_evaluation('Ensemble Method - Voting',predicted_Vote)\n",
    "print('Training Accuracy : %.4f    ||   Training Time  %4.2f seconds' % (votingCLF.score(X_train, y_train),run_time))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67f5b036",
   "metadata": {},
   "source": [
    "#### Bagging  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "9aa52f2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Doing 2-fold Cross Validation Here :\n",
      "Accuracy :0.627147  (+/- 0.0008)  [                 SVC]   ||  42.62 seconds\n",
      "Accuracy :0.592722  (+/- 0.0013)  [                 MNB]   ||  0.02 seconds\n",
      "Accuracy :0.665102  (+/- 0.0028)  [  Random Forest Tree]   ||  3.65 seconds\n",
      "Accuracy :0.683333  (+/- 0.0010)  [ Logistic Regression]   ||  0.22 seconds\n",
      "Accuracy :0.657351  (+/- 0.0013)  [BaggingClassifier(base_estimator=RandomForestClassifier(max_depth=100,\n",
      "                                                        max_leaf_nodes=65,\n",
      "                                                        random_state=1),\n",
      "                  max_samples=0.7, n_estimators=20, oob_score=True,\n",
      "                  random_state=1)]   ||  53.68 seconds\n",
      "\t\t\t Bagging\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ensemble Method - Bagging Classification Report \n",
      "======================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          -1       0.80      0.16      0.26       126\n",
      "           0       0.29      0.88      0.44        77\n",
      "           1       0.76      0.36      0.48       107\n",
      "\n",
      "    accuracy                           0.41       310\n",
      "   macro avg       0.62      0.47      0.39       310\n",
      "weighted avg       0.66      0.41      0.38       310\n",
      "\n",
      "Training Accuracy : 0.6731   ||   Training Time  35.58 seconds\n"
     ]
    }
   ],
   "source": [
    "bag_model=BaggingClassifier(\n",
    "    base_estimator=RandomForestClassifier(random_state=1,max_leaf_nodes=65,max_depth=100),\n",
    "    n_estimators=20, #n_estimator - how many subgroup of dataset\n",
    "    max_samples=0.7, #Use 80%of sample\n",
    "    oob_score=True, #Out of bag \n",
    "    random_state=1\n",
    ")\n",
    "model_accuracy(m1,m2,m3,m4,bag_model) \n",
    "start_time = time.time()\n",
    "bag_model.fit(X_train,y_train)\n",
    "run_time = time.time() - start_time\n",
    "predictedBag=bag_model.predict(X_test)\n",
    "score_Bag=accuracy_score(predictedBag,y_test)\n",
    "\n",
    "print('\\t\\t\\t Bagging')\n",
    "print('================================================================\\n')\n",
    "confusion_matrix_evaluation('Ensemble Method - Bagging',predictedBag)\n",
    "print('Training Accuracy : %.4f   ||   Training Time  %4.2f seconds' % (bag_model.score(X_train, y_train),run_time))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5bbad8b1",
   "metadata": {},
   "source": [
    "### Stacking "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "c1dcc68d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Doing 2-fold Cross Validation Here :\n",
      "Accuracy :0.627147  (+/- 0.0008)  [                 SVC]   ||  38.04 seconds\n",
      "Accuracy :0.592722  (+/- 0.0013)  [                 MNB]   ||  0.01 seconds\n",
      "Accuracy :0.665102  (+/- 0.0028)  [  Random Forest Tree]   ||  3.46 seconds\n",
      "Accuracy :0.683333  (+/- 0.0010)  [ Logistic Regression]   ||  0.37 seconds\n",
      "Accuracy :0.696361  (+/- 0.0033)  [StackingClassifier(estimators=[('SVC',\n",
      "                                SVC(C=0.1, kernel='linear', random_state=1)),\n",
      "                               ('Multinomial Naive Bayes',\n",
      "                                MultinomialNB(alpha=20)),\n",
      "                               ('Random Forest Tree',\n",
      "                                RandomForestClassifier(max_depth=100,\n",
      "                                                       max_leaf_nodes=65,\n",
      "                                                       random_state=1)),\n",
      "                               ('Logistic Regression',\n",
      "                                LogisticRegression(multi_class='multinomial',\n",
      "                                                   random_state=1,\n",
      "                                                   solver='saga'))])]   ||  165.26 seconds\n",
      "\t\t\t Stacking\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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ensemble Method - Stacking Classification Report \n",
      "======================================================\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          -1       0.78      0.25      0.38       126\n",
      "           0       0.28      0.79      0.42        77\n",
      "           1       0.70      0.36      0.47       107\n",
      "\n",
      "    accuracy                           0.42       310\n",
      "   macro avg       0.59      0.47      0.42       310\n",
      "weighted avg       0.63      0.42      0.42       310\n",
      "\n",
      "Training Accuracy : 0.7211   ||   Training Time  238.68 seconds\n"
     ]
    }
   ],
   "source": [
    "stackingCLF=StackingClassifier([('SVC',m1),('Multinomial Naive Bayes',m2),('Random Forest Tree',m3),('Logistic Regression',m4)\n",
    "                       ])\n",
    "model_accuracy(m1,m2,m3,m4,stackingCLF)   \n",
    "start_time = time.time()\n",
    "stackingCLF.fit(X_train,y_train)\n",
    "run_time = time.time() - start_time\n",
    "stack_predicted=stackingCLF.predict(X_test)\n",
    "\n",
    "print('\\t\\t\\t Stacking')\n",
    "print('================================================================\\n')\n",
    "confusion_matrix_evaluation('Ensemble Method - Stacking',stack_predicted)\n",
    "print('Training Accuracy : %.4f   ||   Training Time  %4.2f seconds' % (stackingCLF.score(X_train, y_train),run_time))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}