├── AirBNB_EDA.ipynb
├── Airlines_Passenger_Traffic_TSA.ipynb
├── Animated_Weather_Graphs.ipynb
├── Bangladesh_Economic_Indicators_Analysis.ipynb
├── Bank_Customer_Churn_Prediction.ipynb
├── COVID_19_Analysis_and_Prediction.ipynb
├── Car_Features_Analysis.ipynb
├── Credit_Card_Fraud_Detection.ipynb
├── HR_Analysis.ipynb
├── Heart_Attack_Prediction.ipynb
├── Hotel_Reviews_Sentiment_Prediction.ipynb
├── House_Price_Prediction.ipynb
├── IMDb_Sentiment_Analysis.ipynb
├── IPL_EDA.ipynb
├── Image_Segmentation_using_FastAI.ipynb
├── Iris_Classifier_Model_Comparison.ipynb
├── Loan_Default_Prediction.ipynb
├── Mall_Customers_CLustering.ipynb
├── Movie_Recommendation_Engine.ipynb
├── README.md
├── SMS_Spam_Detection_NLP_.ipynb
├── Stock_Market_Analysis_and_Prediction_.ipynb
├── Suicide_Rate_Analysis.ipynb
├── Udemy_Courses_Recommendation_and_Auditor.ipynb
├── Video_Games_Sale_Prediction_and_EDA.ipynb
└── World_Happiness_EDA.ipynb


/README.md:
--------------------------------------------------------------------------------
 1 | # Data-Analysis-
 2 | Different types of data analytics projects : EDA, PDA, DDA, TSA and much more.....
 3 | ## Contents
 4 | 
 5 | <hr>
 6 | 
 7 | <ol>
 8 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Car_Features_Analysis.ipynb"> Car Feature Analysis</a></li>
 9 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Iris_Classifier_Model_Comparison.ipynb"> Iris Classifier Model Comparison</a></li>
10 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/COVID_19_Analysis_and_Prediction.ipynb">COVID-19 Analysis and Prediction</a></li>
11 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/AirBNB_EDA.ipynb">Airbnb Explanatory Data Analysis</a></li>
12 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/IMDb_Sentiment_Analysis.ipynb">IMDb Sentiment Analysis</a></li>
13 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Mall_Customers_CLustering.ipynb">Mall Customer Clustering</a></li>
14 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/IPL_EDA.ipynb">IPL Explanatory Data Analysis</a></li>
15 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Movie_Recommendation_Engine.ipynb">Movie Recommendation Engine</a></li>
16 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Credit_Card_Fraud_Detection.ipynb">Credit Card Fraud Detection</a></li>
17 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Hotel_Reviews_Sentiment_Prediction.ipynb">Hotel Reviews Sentiment Prediction</a></li>
18 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Loan_Default_Prediction.ipynb">Loan Default Prediction</a></li>
19 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Airlines_Passenger_Traffic_TSA.ipynb">Flight Passenger Traffic : TSA</a></li>
20 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Suicide_Rate_Analysis.ipynb">Suicide Rate Analysis</a></li>
21 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Video_Games_Sale_Prediction_and_EDA.ipynb">Videogames Sales Prediction and EDA</a></li>
22 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Stock_Market_Analysis_and_Prediction_.ipynb">Stock Market Prediction and Analysis using LSTM</a></li>
23 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Bank_Customer_Churn_Prediction.ipynb">Bank Customer Churn Prediction</a></li>
24 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Heart_Attack_Prediction.ipynb">Heart Attack Prediction</a></li>
25 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/SMS_Spam_Detection_NLP_.ipynb">SMS Spam Detection using NLP</a></li>
26 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/World_Happiness_EDA.ipynb">World Happiness EDA</a></li>
27 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/HR_Analysis.ipynb">HR Data Analytics</a></li>
28 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Udemy_Courses_Recommendation_and_Auditor.ipynb">Udemy Course Recommendation and Auditor</a></li>
29 |   <li><a href="https://github.com/MainakRepositor/Data-Analysis/blob/master/Animated_Weather_Graphs.ipynb">Animated Time Series Graphs</a></li>
30 | </ol>
31 | 


--------------------------------------------------------------------------------
/SMS_Spam_Detection_NLP_.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |   "nbformat": 4,
   3 |   "nbformat_minor": 0,
   4 |   "metadata": {
   5 |     "colab": {
   6 |       "name": "SMS Spam Detection : NLP .ipynb",
   7 |       "provenance": [],
   8 |       "authorship_tag": "ABX9TyNJUawe/r6XNkbjIbffEqyp",
   9 |       "include_colab_link": true
  10 |     },
  11 |     "kernelspec": {
  12 |       "name": "python3",
  13 |       "display_name": "Python 3"
  14 |     }
  15 |   },
  16 |   "cells": [
  17 |     {
  18 |       "cell_type": "markdown",
  19 |       "metadata": {
  20 |         "id": "view-in-github",
  21 |         "colab_type": "text"
  22 |       },
  23 |       "source": [
  24 |         "<a href=\"https://colab.research.google.com/github/MainakRepositor/Data-Analysis/blob/master/SMS_Spam_Detection_NLP_.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
  25 |       ]
  26 |     },
  27 |     {
  28 |       "cell_type": "markdown",
  29 |       "metadata": {
  30 |         "id": "EsyQyOF1ccef"
  31 |       },
  32 |       "source": [
  33 |         "# SMS Spam Detection using Natural Language Processing\n",
  34 |         "\n",
  35 |         "<hr>"
  36 |       ]
  37 |     },
  38 |     {
  39 |       "cell_type": "markdown",
  40 |       "metadata": {
  41 |         "id": "vRRCtuLXcqPm"
  42 |       },
  43 |       "source": [
  44 |         "## Approach :\n",
  45 |         "### i.  Representing text as numerical data\n",
  46 |         "### ii. Reading a text-based dataset into pandas\n",
  47 |         "### iii.Vectorizing our dataset\n",
  48 |         "### iv. Building and evaluating a model\n",
  49 |         "### v.  Comparing models\n",
  50 |         "### vi. Examining a model for further insight\n"
  51 |       ]
  52 |     },
  53 |     {
  54 |       "cell_type": "markdown",
  55 |       "metadata": {
  56 |         "id": "QH0PTBSFdBQe"
  57 |       },
  58 |       "source": [
  59 |         "## Importing necessary libraries"
  60 |       ]
  61 |     },
  62 |     {
  63 |       "cell_type": "code",
  64 |       "metadata": {
  65 |         "colab": {
  66 |           "base_uri": "https://localhost:8080/"
  67 |         },
  68 |         "id": "wSJGpEu1cWmI",
  69 |         "outputId": "07098125-2340-48cf-a0cd-67440e9f24db"
  70 |       },
  71 |       "source": [
  72 |         "import pandas as pd\n",
  73 |         "import numpy as np\n",
  74 |         "import matplotlib.pyplot as plt\n",
  75 |         "import seaborn as sns\n",
  76 |         "import nltk\n",
  77 |         "nltk.download('stopwords')\n",
  78 |         "%matplotlib inline\n",
  79 |         "sns.set_style(\"darkgrid\")\n",
  80 |         "plt.style.use(\"fivethirtyeight\")\n",
  81 |         "print(\"Necessary packages included successfully!\")"
  82 |       ],
  83 |       "execution_count": 27,
  84 |       "outputs": [
  85 |         {
  86 |           "output_type": "stream",
  87 |           "text": [
  88 |             "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
  89 |             "[nltk_data]   Unzipping corpora/stopwords.zip.\n",
  90 |             "Necessary packages included successfully!\n"
  91 |           ],
  92 |           "name": "stdout"
  93 |         }
  94 |       ]
  95 |     },
  96 |     {
  97 |       "cell_type": "markdown",
  98 |       "metadata": {
  99 |         "id": "6gz7JglLjH8U"
 100 |       },
 101 |       "source": [
 102 |         "## 1. Representing text as numerical data"
 103 |       ]
 104 |     },
 105 |     {
 106 |       "cell_type": "code",
 107 |       "metadata": {
 108 |         "id": "HEh2dqT6i7-i"
 109 |       },
 110 |       "source": [
 111 |         "# example text for model training (SMS messages)\n",
 112 |         "simple_train = ['call you tonight', 'Call me a cab', 'Please call me... PLEASE!']"
 113 |       ],
 114 |       "execution_count": 3,
 115 |       "outputs": []
 116 |     },
 117 |     {
 118 |       "cell_type": "code",
 119 |       "metadata": {
 120 |         "id": "QhGclG-ZjRui"
 121 |       },
 122 |       "source": [
 123 |         "# import and instantiate CountVectorizer (with the default parameters)\n",
 124 |         "from sklearn.feature_extraction.text import CountVectorizer\n",
 125 |         "vect = CountVectorizer()"
 126 |       ],
 127 |       "execution_count": 4,
 128 |       "outputs": []
 129 |     },
 130 |     {
 131 |       "cell_type": "code",
 132 |       "metadata": {
 133 |         "colab": {
 134 |           "base_uri": "https://localhost:8080/"
 135 |         },
 136 |         "id": "2OZQeUtxj2a3",
 137 |         "outputId": "8e0bd715-c0c1-4587-a4a0-1865dcaadd56"
 138 |       },
 139 |       "source": [
 140 |         "# learn the 'vocabulary' of the training data (occurs in-place)\n",
 141 |         "vect.fit(simple_train)"
 142 |       ],
 143 |       "execution_count": 5,
 144 |       "outputs": [
 145 |         {
 146 |           "output_type": "execute_result",
 147 |           "data": {
 148 |             "text/plain": [
 149 |               "CountVectorizer(analyzer='word', binary=False, decode_error='strict',\n",
 150 |               "                dtype=<class 'numpy.int64'>, encoding='utf-8', input='content',\n",
 151 |               "                lowercase=True, max_df=1.0, max_features=None, min_df=1,\n",
 152 |               "                ngram_range=(1, 1), preprocessor=None, stop_words=None,\n",
 153 |               "                strip_accents=None, token_pattern='(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
 154 |               "                tokenizer=None, vocabulary=None)"
 155 |             ]
 156 |           },
 157 |           "metadata": {
 158 |             "tags": []
 159 |           },
 160 |           "execution_count": 5
 161 |         }
 162 |       ]
 163 |     },
 164 |     {
 165 |       "cell_type": "code",
 166 |       "metadata": {
 167 |         "colab": {
 168 |           "base_uri": "https://localhost:8080/"
 169 |         },
 170 |         "id": "3mwQZpkFj5Hw",
 171 |         "outputId": "981e5255-4d28-44dd-e1d0-4fb5d809b713"
 172 |       },
 173 |       "source": [
 174 |         "# examine the fitted vocabulary\n",
 175 |         "vect.get_feature_names()"
 176 |       ],
 177 |       "execution_count": 6,
 178 |       "outputs": [
 179 |         {
 180 |           "output_type": "execute_result",
 181 |           "data": {
 182 |             "text/plain": [
 183 |               "['cab', 'call', 'me', 'please', 'tonight', 'you']"
 184 |             ]
 185 |           },
 186 |           "metadata": {
 187 |             "tags": []
 188 |           },
 189 |           "execution_count": 6
 190 |         }
 191 |       ]
 192 |     },
 193 |     {
 194 |       "cell_type": "code",
 195 |       "metadata": {
 196 |         "colab": {
 197 |           "base_uri": "https://localhost:8080/"
 198 |         },
 199 |         "id": "DYXT6QnCj8Y-",
 200 |         "outputId": "e0df455a-6810-4298-8396-030a1aa58cfb"
 201 |       },
 202 |       "source": [
 203 |         "# transform training data into a 'document-term matrix'\n",
 204 |         "simple_train_dtm = vect.transform(simple_train)\n",
 205 |         "simple_train_dtm"
 206 |       ],
 207 |       "execution_count": 7,
 208 |       "outputs": [
 209 |         {
 210 |           "output_type": "execute_result",
 211 |           "data": {
 212 |             "text/plain": [
 213 |               "<3x6 sparse matrix of type '<class 'numpy.int64'>'\n",
 214 |               "\twith 9 stored elements in Compressed Sparse Row format>"
 215 |             ]
 216 |           },
 217 |           "metadata": {
 218 |             "tags": []
 219 |           },
 220 |           "execution_count": 7
 221 |         }
 222 |       ]
 223 |     },
 224 |     {
 225 |       "cell_type": "code",
 226 |       "metadata": {
 227 |         "colab": {
 228 |           "base_uri": "https://localhost:8080/"
 229 |         },
 230 |         "id": "Vx0TufUWj_f7",
 231 |         "outputId": "d7851980-c53a-41a7-9ebb-aad6b7b9bf7d"
 232 |       },
 233 |       "source": [
 234 |         "# convert sparse matrix to a dense matrix\n",
 235 |         "simple_train_dtm.toarray()"
 236 |       ],
 237 |       "execution_count": 8,
 238 |       "outputs": [
 239 |         {
 240 |           "output_type": "execute_result",
 241 |           "data": {
 242 |             "text/plain": [
 243 |               "array([[0, 1, 0, 0, 1, 1],\n",
 244 |               "       [1, 1, 1, 0, 0, 0],\n",
 245 |               "       [0, 1, 1, 2, 0, 0]])"
 246 |             ]
 247 |           },
 248 |           "metadata": {
 249 |             "tags": []
 250 |           },
 251 |           "execution_count": 8
 252 |         }
 253 |       ]
 254 |     },
 255 |     {
 256 |       "cell_type": "code",
 257 |       "metadata": {
 258 |         "colab": {
 259 |           "base_uri": "https://localhost:8080/",
 260 |           "height": 137
 261 |         },
 262 |         "id": "L72IJthbkDH-",
 263 |         "outputId": "5e34feb8-9219-4c14-a5b9-16d4fd626c25"
 264 |       },
 265 |       "source": [
 266 |         "# examine the vocabulary and document-term matrix together\n",
 267 |         "pd.DataFrame(simple_train_dtm.toarray(), columns=vect.get_feature_names())"
 268 |       ],
 269 |       "execution_count": 9,
 270 |       "outputs": [
 271 |         {
 272 |           "output_type": "execute_result",
 273 |           "data": {
 274 |             "text/html": [
 275 |               "<div>\n",
 276 |               "<style scoped>\n",
 277 |               "    .dataframe tbody tr th:only-of-type {\n",
 278 |               "        vertical-align: middle;\n",
 279 |               "    }\n",
 280 |               "\n",
 281 |               "    .dataframe tbody tr th {\n",
 282 |               "        vertical-align: top;\n",
 283 |               "    }\n",
 284 |               "\n",
 285 |               "    .dataframe thead th {\n",
 286 |               "        text-align: right;\n",
 287 |               "    }\n",
 288 |               "</style>\n",
 289 |               "<table border=\"1\" class=\"dataframe\">\n",
 290 |               "  <thead>\n",
 291 |               "    <tr style=\"text-align: right;\">\n",
 292 |               "      <th></th>\n",
 293 |               "      <th>cab</th>\n",
 294 |               "      <th>call</th>\n",
 295 |               "      <th>me</th>\n",
 296 |               "      <th>please</th>\n",
 297 |               "      <th>tonight</th>\n",
 298 |               "      <th>you</th>\n",
 299 |               "    </tr>\n",
 300 |               "  </thead>\n",
 301 |               "  <tbody>\n",
 302 |               "    <tr>\n",
 303 |               "      <th>0</th>\n",
 304 |               "      <td>0</td>\n",
 305 |               "      <td>1</td>\n",
 306 |               "      <td>0</td>\n",
 307 |               "      <td>0</td>\n",
 308 |               "      <td>1</td>\n",
 309 |               "      <td>1</td>\n",
 310 |               "    </tr>\n",
 311 |               "    <tr>\n",
 312 |               "      <th>1</th>\n",
 313 |               "      <td>1</td>\n",
 314 |               "      <td>1</td>\n",
 315 |               "      <td>1</td>\n",
 316 |               "      <td>0</td>\n",
 317 |               "      <td>0</td>\n",
 318 |               "      <td>0</td>\n",
 319 |               "    </tr>\n",
 320 |               "    <tr>\n",
 321 |               "      <th>2</th>\n",
 322 |               "      <td>0</td>\n",
 323 |               "      <td>1</td>\n",
 324 |               "      <td>1</td>\n",
 325 |               "      <td>2</td>\n",
 326 |               "      <td>0</td>\n",
 327 |               "      <td>0</td>\n",
 328 |               "    </tr>\n",
 329 |               "  </tbody>\n",
 330 |               "</table>\n",
 331 |               "</div>"
 332 |             ],
 333 |             "text/plain": [
 334 |               "   cab  call  me  please  tonight  you\n",
 335 |               "0    0     1   0       0        1    1\n",
 336 |               "1    1     1   1       0        0    0\n",
 337 |               "2    0     1   1       2        0    0"
 338 |             ]
 339 |           },
 340 |           "metadata": {
 341 |             "tags": []
 342 |           },
 343 |           "execution_count": 9
 344 |         }
 345 |       ]
 346 |     },
 347 |     {
 348 |       "cell_type": "code",
 349 |       "metadata": {
 350 |         "colab": {
 351 |           "base_uri": "https://localhost:8080/"
 352 |         },
 353 |         "id": "__82s5G9kGSb",
 354 |         "outputId": "d7457d64-a598-43dd-a1ed-e88915271b98"
 355 |       },
 356 |       "source": [
 357 |         "# check the type of the document-term matrix\n",
 358 |         "type(simple_train_dtm)"
 359 |       ],
 360 |       "execution_count": 10,
 361 |       "outputs": [
 362 |         {
 363 |           "output_type": "execute_result",
 364 |           "data": {
 365 |             "text/plain": [
 366 |               "scipy.sparse.csr.csr_matrix"
 367 |             ]
 368 |           },
 369 |           "metadata": {
 370 |             "tags": []
 371 |           },
 372 |           "execution_count": 10
 373 |         }
 374 |       ]
 375 |     },
 376 |     {
 377 |       "cell_type": "code",
 378 |       "metadata": {
 379 |         "colab": {
 380 |           "base_uri": "https://localhost:8080/"
 381 |         },
 382 |         "id": "HGBIqef6kLPj",
 383 |         "outputId": "2078557e-bd13-4485-a2fe-590797e10422"
 384 |       },
 385 |       "source": [
 386 |         "# examine the sparse matrix contents\n",
 387 |         "print(simple_train_dtm)"
 388 |       ],
 389 |       "execution_count": 11,
 390 |       "outputs": [
 391 |         {
 392 |           "output_type": "stream",
 393 |           "text": [
 394 |             "  (0, 1)\t1\n",
 395 |             "  (0, 4)\t1\n",
 396 |             "  (0, 5)\t1\n",
 397 |             "  (1, 0)\t1\n",
 398 |             "  (1, 1)\t1\n",
 399 |             "  (1, 2)\t1\n",
 400 |             "  (2, 1)\t1\n",
 401 |             "  (2, 2)\t1\n",
 402 |             "  (2, 3)\t2\n"
 403 |           ],
 404 |           "name": "stdout"
 405 |         }
 406 |       ]
 407 |     },
 408 |     {
 409 |       "cell_type": "code",
 410 |       "metadata": {
 411 |         "id": "kMv7gIQ4kNx8"
 412 |       },
 413 |       "source": [
 414 |         "# example text for model testing\n",
 415 |         "simple_test = [\"please don't call me\"]"
 416 |       ],
 417 |       "execution_count": 12,
 418 |       "outputs": []
 419 |     },
 420 |     {
 421 |       "cell_type": "code",
 422 |       "metadata": {
 423 |         "colab": {
 424 |           "base_uri": "https://localhost:8080/"
 425 |         },
 426 |         "id": "v76GaavFkSDd",
 427 |         "outputId": "f4c4e694-1100-48ee-c888-2fade4a7b330"
 428 |       },
 429 |       "source": [
 430 |         "# transform testing data into a document-term matrix (using existing vocabulary)\n",
 431 |         "simple_test_dtm = vect.transform(simple_test)\n",
 432 |         "simple_test_dtm.toarray()"
 433 |       ],
 434 |       "execution_count": 13,
 435 |       "outputs": [
 436 |         {
 437 |           "output_type": "execute_result",
 438 |           "data": {
 439 |             "text/plain": [
 440 |               "array([[0, 1, 1, 1, 0, 0]])"
 441 |             ]
 442 |           },
 443 |           "metadata": {
 444 |             "tags": []
 445 |           },
 446 |           "execution_count": 13
 447 |         }
 448 |       ]
 449 |     },
 450 |     {
 451 |       "cell_type": "code",
 452 |       "metadata": {
 453 |         "colab": {
 454 |           "base_uri": "https://localhost:8080/",
 455 |           "height": 77
 456 |         },
 457 |         "id": "jOZOL10OkUh6",
 458 |         "outputId": "62dffe55-9741-4b14-ba86-aa4faab32681"
 459 |       },
 460 |       "source": [
 461 |         "# examine the vocabulary and document-term matrix together\n",
 462 |         "pd.DataFrame(simple_test_dtm.toarray(), columns=vect.get_feature_names())"
 463 |       ],
 464 |       "execution_count": 14,
 465 |       "outputs": [
 466 |         {
 467 |           "output_type": "execute_result",
 468 |           "data": {
 469 |             "text/html": [
 470 |               "<div>\n",
 471 |               "<style scoped>\n",
 472 |               "    .dataframe tbody tr th:only-of-type {\n",
 473 |               "        vertical-align: middle;\n",
 474 |               "    }\n",
 475 |               "\n",
 476 |               "    .dataframe tbody tr th {\n",
 477 |               "        vertical-align: top;\n",
 478 |               "    }\n",
 479 |               "\n",
 480 |               "    .dataframe thead th {\n",
 481 |               "        text-align: right;\n",
 482 |               "    }\n",
 483 |               "</style>\n",
 484 |               "<table border=\"1\" class=\"dataframe\">\n",
 485 |               "  <thead>\n",
 486 |               "    <tr style=\"text-align: right;\">\n",
 487 |               "      <th></th>\n",
 488 |               "      <th>cab</th>\n",
 489 |               "      <th>call</th>\n",
 490 |               "      <th>me</th>\n",
 491 |               "      <th>please</th>\n",
 492 |               "      <th>tonight</th>\n",
 493 |               "      <th>you</th>\n",
 494 |               "    </tr>\n",
 495 |               "  </thead>\n",
 496 |               "  <tbody>\n",
 497 |               "    <tr>\n",
 498 |               "      <th>0</th>\n",
 499 |               "      <td>0</td>\n",
 500 |               "      <td>1</td>\n",
 501 |               "      <td>1</td>\n",
 502 |               "      <td>1</td>\n",
 503 |               "      <td>0</td>\n",
 504 |               "      <td>0</td>\n",
 505 |               "    </tr>\n",
 506 |               "  </tbody>\n",
 507 |               "</table>\n",
 508 |               "</div>"
 509 |             ],
 510 |             "text/plain": [
 511 |               "   cab  call  me  please  tonight  you\n",
 512 |               "0    0     1   1       1        0    0"
 513 |             ]
 514 |           },
 515 |           "metadata": {
 516 |             "tags": []
 517 |           },
 518 |           "execution_count": 14
 519 |         }
 520 |       ]
 521 |     },
 522 |     {
 523 |       "cell_type": "markdown",
 524 |       "metadata": {
 525 |         "id": "ghaGexa1kwEQ"
 526 |       },
 527 |       "source": [
 528 |         "## 2. Reading a text-based dataset into notebook"
 529 |       ]
 530 |     },
 531 |     {
 532 |       "cell_type": "code",
 533 |       "metadata": {
 534 |         "colab": {
 535 |           "base_uri": "https://localhost:8080/",
 536 |           "height": 196
 537 |         },
 538 |         "id": "Tie7p7IfkXpQ",
 539 |         "outputId": "5d01f03c-0429-4e0e-a049-1c4a4b7bc7a9"
 540 |       },
 541 |       "source": [
 542 |         "url = 'https://raw.githubusercontent.com/MainakRepositor/Datasets-/master/spam.csv'\n",
 543 |         "sms = pd.read_csv(url,error_bad_lines=False,encoding='latin-1')\n",
 544 |         "sms.dropna(how=\"any\", inplace=True, axis=1)\n",
 545 |         "sms.columns = ['label', 'message']\n",
 546 |         "sms.head()"
 547 |       ],
 548 |       "execution_count": 15,
 549 |       "outputs": [
 550 |         {
 551 |           "output_type": "execute_result",
 552 |           "data": {
 553 |             "text/html": [
 554 |               "<div>\n",
 555 |               "<style scoped>\n",
 556 |               "    .dataframe tbody tr th:only-of-type {\n",
 557 |               "        vertical-align: middle;\n",
 558 |               "    }\n",
 559 |               "\n",
 560 |               "    .dataframe tbody tr th {\n",
 561 |               "        vertical-align: top;\n",
 562 |               "    }\n",
 563 |               "\n",
 564 |               "    .dataframe thead th {\n",
 565 |               "        text-align: right;\n",
 566 |               "    }\n",
 567 |               "</style>\n",
 568 |               "<table border=\"1\" class=\"dataframe\">\n",
 569 |               "  <thead>\n",
 570 |               "    <tr style=\"text-align: right;\">\n",
 571 |               "      <th></th>\n",
 572 |               "      <th>label</th>\n",
 573 |               "      <th>message</th>\n",
 574 |               "    </tr>\n",
 575 |               "  </thead>\n",
 576 |               "  <tbody>\n",
 577 |               "    <tr>\n",
 578 |               "      <th>0</th>\n",
 579 |               "      <td>ham</td>\n",
 580 |               "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
 581 |               "    </tr>\n",
 582 |               "    <tr>\n",
 583 |               "      <th>1</th>\n",
 584 |               "      <td>ham</td>\n",
 585 |               "      <td>Ok lar... Joking wif u oni...</td>\n",
 586 |               "    </tr>\n",
 587 |               "    <tr>\n",
 588 |               "      <th>2</th>\n",
 589 |               "      <td>spam</td>\n",
 590 |               "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
 591 |               "    </tr>\n",
 592 |               "    <tr>\n",
 593 |               "      <th>3</th>\n",
 594 |               "      <td>ham</td>\n",
 595 |               "      <td>U dun say so early hor... U c already then say...</td>\n",
 596 |               "    </tr>\n",
 597 |               "    <tr>\n",
 598 |               "      <th>4</th>\n",
 599 |               "      <td>ham</td>\n",
 600 |               "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
 601 |               "    </tr>\n",
 602 |               "  </tbody>\n",
 603 |               "</table>\n",
 604 |               "</div>"
 605 |             ],
 606 |             "text/plain": [
 607 |               "  label                                            message\n",
 608 |               "0   ham  Go until jurong point, crazy.. Available only ...\n",
 609 |               "1   ham                      Ok lar... Joking wif u oni...\n",
 610 |               "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
 611 |               "3   ham  U dun say so early hor... U c already then say...\n",
 612 |               "4   ham  Nah I don't think he goes to usf, he lives aro..."
 613 |             ]
 614 |           },
 615 |           "metadata": {
 616 |             "tags": []
 617 |           },
 618 |           "execution_count": 15
 619 |         }
 620 |       ]
 621 |     },
 622 |     {
 623 |       "cell_type": "markdown",
 624 |       "metadata": {
 625 |         "id": "ftTUkK67k0Wd"
 626 |       },
 627 |       "source": [
 628 |         "## 3. Exploratory Data Analysis (EDA)"
 629 |       ]
 630 |     },
 631 |     {
 632 |       "cell_type": "code",
 633 |       "metadata": {
 634 |         "colab": {
 635 |           "base_uri": "https://localhost:8080/",
 636 |           "height": 166
 637 |         },
 638 |         "id": "tgkEiFG9kqrl",
 639 |         "outputId": "4971ab2d-026a-4a91-c5c2-d68e91de6586"
 640 |       },
 641 |       "source": [
 642 |         "sms.describe()"
 643 |       ],
 644 |       "execution_count": 16,
 645 |       "outputs": [
 646 |         {
 647 |           "output_type": "execute_result",
 648 |           "data": {
 649 |             "text/html": [
 650 |               "<div>\n",
 651 |               "<style scoped>\n",
 652 |               "    .dataframe tbody tr th:only-of-type {\n",
 653 |               "        vertical-align: middle;\n",
 654 |               "    }\n",
 655 |               "\n",
 656 |               "    .dataframe tbody tr th {\n",
 657 |               "        vertical-align: top;\n",
 658 |               "    }\n",
 659 |               "\n",
 660 |               "    .dataframe thead th {\n",
 661 |               "        text-align: right;\n",
 662 |               "    }\n",
 663 |               "</style>\n",
 664 |               "<table border=\"1\" class=\"dataframe\">\n",
 665 |               "  <thead>\n",
 666 |               "    <tr style=\"text-align: right;\">\n",
 667 |               "      <th></th>\n",
 668 |               "      <th>label</th>\n",
 669 |               "      <th>message</th>\n",
 670 |               "    </tr>\n",
 671 |               "  </thead>\n",
 672 |               "  <tbody>\n",
 673 |               "    <tr>\n",
 674 |               "      <th>count</th>\n",
 675 |               "      <td>5572</td>\n",
 676 |               "      <td>5572</td>\n",
 677 |               "    </tr>\n",
 678 |               "    <tr>\n",
 679 |               "      <th>unique</th>\n",
 680 |               "      <td>2</td>\n",
 681 |               "      <td>5169</td>\n",
 682 |               "    </tr>\n",
 683 |               "    <tr>\n",
 684 |               "      <th>top</th>\n",
 685 |               "      <td>ham</td>\n",
 686 |               "      <td>Sorry, I'll call later</td>\n",
 687 |               "    </tr>\n",
 688 |               "    <tr>\n",
 689 |               "      <th>freq</th>\n",
 690 |               "      <td>4825</td>\n",
 691 |               "      <td>30</td>\n",
 692 |               "    </tr>\n",
 693 |               "  </tbody>\n",
 694 |               "</table>\n",
 695 |               "</div>"
 696 |             ],
 697 |             "text/plain": [
 698 |               "       label                 message\n",
 699 |               "count   5572                    5572\n",
 700 |               "unique     2                    5169\n",
 701 |               "top      ham  Sorry, I'll call later\n",
 702 |               "freq    4825                      30"
 703 |             ]
 704 |           },
 705 |           "metadata": {
 706 |             "tags": []
 707 |           },
 708 |           "execution_count": 16
 709 |         }
 710 |       ]
 711 |     },
 712 |     {
 713 |       "cell_type": "code",
 714 |       "metadata": {
 715 |         "colab": {
 716 |           "base_uri": "https://localhost:8080/",
 717 |           "height": 166
 718 |         },
 719 |         "id": "8tOf8rdrk5pu",
 720 |         "outputId": "8831c62e-7b70-4be0-a33f-5ae27015acbe"
 721 |       },
 722 |       "source": [
 723 |         "sms.groupby('label').describe()"
 724 |       ],
 725 |       "execution_count": 17,
 726 |       "outputs": [
 727 |         {
 728 |           "output_type": "execute_result",
 729 |           "data": {
 730 |             "text/html": [
 731 |               "<div>\n",
 732 |               "<style scoped>\n",
 733 |               "    .dataframe tbody tr th:only-of-type {\n",
 734 |               "        vertical-align: middle;\n",
 735 |               "    }\n",
 736 |               "\n",
 737 |               "    .dataframe tbody tr th {\n",
 738 |               "        vertical-align: top;\n",
 739 |               "    }\n",
 740 |               "\n",
 741 |               "    .dataframe thead tr th {\n",
 742 |               "        text-align: left;\n",
 743 |               "    }\n",
 744 |               "\n",
 745 |               "    .dataframe thead tr:last-of-type th {\n",
 746 |               "        text-align: right;\n",
 747 |               "    }\n",
 748 |               "</style>\n",
 749 |               "<table border=\"1\" class=\"dataframe\">\n",
 750 |               "  <thead>\n",
 751 |               "    <tr>\n",
 752 |               "      <th></th>\n",
 753 |               "      <th colspan=\"4\" halign=\"left\">message</th>\n",
 754 |               "    </tr>\n",
 755 |               "    <tr>\n",
 756 |               "      <th></th>\n",
 757 |               "      <th>count</th>\n",
 758 |               "      <th>unique</th>\n",
 759 |               "      <th>top</th>\n",
 760 |               "      <th>freq</th>\n",
 761 |               "    </tr>\n",
 762 |               "    <tr>\n",
 763 |               "      <th>label</th>\n",
 764 |               "      <th></th>\n",
 765 |               "      <th></th>\n",
 766 |               "      <th></th>\n",
 767 |               "      <th></th>\n",
 768 |               "    </tr>\n",
 769 |               "  </thead>\n",
 770 |               "  <tbody>\n",
 771 |               "    <tr>\n",
 772 |               "      <th>ham</th>\n",
 773 |               "      <td>4825</td>\n",
 774 |               "      <td>4516</td>\n",
 775 |               "      <td>Sorry, I'll call later</td>\n",
 776 |               "      <td>30</td>\n",
 777 |               "    </tr>\n",
 778 |               "    <tr>\n",
 779 |               "      <th>spam</th>\n",
 780 |               "      <td>747</td>\n",
 781 |               "      <td>653</td>\n",
 782 |               "      <td>Please call our customer service representativ...</td>\n",
 783 |               "      <td>4</td>\n",
 784 |               "    </tr>\n",
 785 |               "  </tbody>\n",
 786 |               "</table>\n",
 787 |               "</div>"
 788 |             ],
 789 |             "text/plain": [
 790 |               "      message                                                               \n",
 791 |               "        count unique                                                top freq\n",
 792 |               "label                                                                       \n",
 793 |               "ham      4825   4516                             Sorry, I'll call later   30\n",
 794 |               "spam      747    653  Please call our customer service representativ...    4"
 795 |             ]
 796 |           },
 797 |           "metadata": {
 798 |             "tags": []
 799 |           },
 800 |           "execution_count": 17
 801 |         }
 802 |       ]
 803 |     },
 804 |     {
 805 |       "cell_type": "code",
 806 |       "metadata": {
 807 |         "colab": {
 808 |           "base_uri": "https://localhost:8080/",
 809 |           "height": 196
 810 |         },
 811 |         "id": "fvkzXgIxk8qd",
 812 |         "outputId": "a39c1b97-5d1b-4196-ae9f-222389039515"
 813 |       },
 814 |       "source": [
 815 |         "# convert label to a numerical variable\n",
 816 |         "sms['label_num'] = sms.label.map({'ham':0, 'spam':1})\n",
 817 |         "sms.head()"
 818 |       ],
 819 |       "execution_count": 18,
 820 |       "outputs": [
 821 |         {
 822 |           "output_type": "execute_result",
 823 |           "data": {
 824 |             "text/html": [
 825 |               "<div>\n",
 826 |               "<style scoped>\n",
 827 |               "    .dataframe tbody tr th:only-of-type {\n",
 828 |               "        vertical-align: middle;\n",
 829 |               "    }\n",
 830 |               "\n",
 831 |               "    .dataframe tbody tr th {\n",
 832 |               "        vertical-align: top;\n",
 833 |               "    }\n",
 834 |               "\n",
 835 |               "    .dataframe thead th {\n",
 836 |               "        text-align: right;\n",
 837 |               "    }\n",
 838 |               "</style>\n",
 839 |               "<table border=\"1\" class=\"dataframe\">\n",
 840 |               "  <thead>\n",
 841 |               "    <tr style=\"text-align: right;\">\n",
 842 |               "      <th></th>\n",
 843 |               "      <th>label</th>\n",
 844 |               "      <th>message</th>\n",
 845 |               "      <th>label_num</th>\n",
 846 |               "    </tr>\n",
 847 |               "  </thead>\n",
 848 |               "  <tbody>\n",
 849 |               "    <tr>\n",
 850 |               "      <th>0</th>\n",
 851 |               "      <td>ham</td>\n",
 852 |               "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
 853 |               "      <td>0</td>\n",
 854 |               "    </tr>\n",
 855 |               "    <tr>\n",
 856 |               "      <th>1</th>\n",
 857 |               "      <td>ham</td>\n",
 858 |               "      <td>Ok lar... Joking wif u oni...</td>\n",
 859 |               "      <td>0</td>\n",
 860 |               "    </tr>\n",
 861 |               "    <tr>\n",
 862 |               "      <th>2</th>\n",
 863 |               "      <td>spam</td>\n",
 864 |               "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
 865 |               "      <td>1</td>\n",
 866 |               "    </tr>\n",
 867 |               "    <tr>\n",
 868 |               "      <th>3</th>\n",
 869 |               "      <td>ham</td>\n",
 870 |               "      <td>U dun say so early hor... U c already then say...</td>\n",
 871 |               "      <td>0</td>\n",
 872 |               "    </tr>\n",
 873 |               "    <tr>\n",
 874 |               "      <th>4</th>\n",
 875 |               "      <td>ham</td>\n",
 876 |               "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
 877 |               "      <td>0</td>\n",
 878 |               "    </tr>\n",
 879 |               "  </tbody>\n",
 880 |               "</table>\n",
 881 |               "</div>"
 882 |             ],
 883 |             "text/plain": [
 884 |               "  label                                            message  label_num\n",
 885 |               "0   ham  Go until jurong point, crazy.. Available only ...          0\n",
 886 |               "1   ham                      Ok lar... Joking wif u oni...          0\n",
 887 |               "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...          1\n",
 888 |               "3   ham  U dun say so early hor... U c already then say...          0\n",
 889 |               "4   ham  Nah I don't think he goes to usf, he lives aro...          0"
 890 |             ]
 891 |           },
 892 |           "metadata": {
 893 |             "tags": []
 894 |           },
 895 |           "execution_count": 18
 896 |         }
 897 |       ]
 898 |     },
 899 |     {
 900 |       "cell_type": "code",
 901 |       "metadata": {
 902 |         "colab": {
 903 |           "base_uri": "https://localhost:8080/",
 904 |           "height": 196
 905 |         },
 906 |         "id": "0tH1qcZ4lAJh",
 907 |         "outputId": "19fe514d-c68f-4a6e-c396-d6b7ec43e95c"
 908 |       },
 909 |       "source": [
 910 |         "sms['message_len'] = sms.message.apply(len)\n",
 911 |         "sms.head()"
 912 |       ],
 913 |       "execution_count": 19,
 914 |       "outputs": [
 915 |         {
 916 |           "output_type": "execute_result",
 917 |           "data": {
 918 |             "text/html": [
 919 |               "<div>\n",
 920 |               "<style scoped>\n",
 921 |               "    .dataframe tbody tr th:only-of-type {\n",
 922 |               "        vertical-align: middle;\n",
 923 |               "    }\n",
 924 |               "\n",
 925 |               "    .dataframe tbody tr th {\n",
 926 |               "        vertical-align: top;\n",
 927 |               "    }\n",
 928 |               "\n",
 929 |               "    .dataframe thead th {\n",
 930 |               "        text-align: right;\n",
 931 |               "    }\n",
 932 |               "</style>\n",
 933 |               "<table border=\"1\" class=\"dataframe\">\n",
 934 |               "  <thead>\n",
 935 |               "    <tr style=\"text-align: right;\">\n",
 936 |               "      <th></th>\n",
 937 |               "      <th>label</th>\n",
 938 |               "      <th>message</th>\n",
 939 |               "      <th>label_num</th>\n",
 940 |               "      <th>message_len</th>\n",
 941 |               "    </tr>\n",
 942 |               "  </thead>\n",
 943 |               "  <tbody>\n",
 944 |               "    <tr>\n",
 945 |               "      <th>0</th>\n",
 946 |               "      <td>ham</td>\n",
 947 |               "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
 948 |               "      <td>0</td>\n",
 949 |               "      <td>111</td>\n",
 950 |               "    </tr>\n",
 951 |               "    <tr>\n",
 952 |               "      <th>1</th>\n",
 953 |               "      <td>ham</td>\n",
 954 |               "      <td>Ok lar... Joking wif u oni...</td>\n",
 955 |               "      <td>0</td>\n",
 956 |               "      <td>29</td>\n",
 957 |               "    </tr>\n",
 958 |               "    <tr>\n",
 959 |               "      <th>2</th>\n",
 960 |               "      <td>spam</td>\n",
 961 |               "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
 962 |               "      <td>1</td>\n",
 963 |               "      <td>155</td>\n",
 964 |               "    </tr>\n",
 965 |               "    <tr>\n",
 966 |               "      <th>3</th>\n",
 967 |               "      <td>ham</td>\n",
 968 |               "      <td>U dun say so early hor... U c already then say...</td>\n",
 969 |               "      <td>0</td>\n",
 970 |               "      <td>49</td>\n",
 971 |               "    </tr>\n",
 972 |               "    <tr>\n",
 973 |               "      <th>4</th>\n",
 974 |               "      <td>ham</td>\n",
 975 |               "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
 976 |               "      <td>0</td>\n",
 977 |               "      <td>61</td>\n",
 978 |               "    </tr>\n",
 979 |               "  </tbody>\n",
 980 |               "</table>\n",
 981 |               "</div>"
 982 |             ],
 983 |             "text/plain": [
 984 |               "  label  ... message_len\n",
 985 |               "0   ham  ...         111\n",
 986 |               "1   ham  ...          29\n",
 987 |               "2  spam  ...         155\n",
 988 |               "3   ham  ...          49\n",
 989 |               "4   ham  ...          61\n",
 990 |               "\n",
 991 |               "[5 rows x 4 columns]"
 992 |             ]
 993 |           },
 994 |           "metadata": {
 995 |             "tags": []
 996 |           },
 997 |           "execution_count": 19
 998 |         }
 999 |       ]
1000 |     },
1001 |     {
1002 |       "cell_type": "code",
1003 |       "metadata": {
1004 |         "colab": {
1005 |           "base_uri": "https://localhost:8080/",
1006 |           "height": 553
1007 |         },
1008 |         "id": "bWwHe6YWlDRa",
1009 |         "outputId": "f50b177e-854b-43f3-9a66-fcdd8752fccd"
1010 |       },
1011 |       "source": [
1012 |         "plt.figure(figsize=(17, 8))\n",
1013 |         "\n",
1014 |         "sms[sms.label=='ham'].message_len.plot(bins=35, kind='hist', color='blue', \n",
1015 |         "                                       label='Ham messages', alpha=0.6)\n",
1016 |         "sms[sms.label=='spam'].message_len.plot(kind='hist', color='red', \n",
1017 |         "                                       label='Spam messages', alpha=0.6)\n",
1018 |         "plt.legend()\n",
1019 |         "plt.xlabel(\"Message Length\")"
1020 |       ],
1021 |       "execution_count": 20,
1022 |       "outputs": [
1023 |         {
1024 |           "output_type": "execute_result",
1025 |           "data": {
1026 |             "text/plain": [
1027 |               "Text(0.5, 0, 'Message Length')"
1028 |             ]
1029 |           },
1030 |           "metadata": {
1031 |             "tags": []
1032 |           },
1033 |           "execution_count": 20
1034 |         },
1035 |         {
1036 |           "output_type": "display_data",
1037 |           "data": {
1038 |             "image/png": "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\n",
1039 |             "text/plain": [
1040 |               "<Figure size 1224x576 with 1 Axes>"
1041 |             ]
1042 |           },
1043 |           "metadata": {
1044 |             "tags": []
1045 |           }
1046 |         }
1047 |       ]
1048 |     },
1049 |     {
1050 |       "cell_type": "code",
1051 |       "metadata": {
1052 |         "colab": {
1053 |           "base_uri": "https://localhost:8080/",
1054 |           "height": 286
1055 |         },
1056 |         "id": "pEax9b3WlHCT",
1057 |         "outputId": "f058c836-36a2-4dd2-9b2a-001ce8540432"
1058 |       },
1059 |       "source": [
1060 |         "sms[sms.label=='ham'].describe()"
1061 |       ],
1062 |       "execution_count": 21,
1063 |       "outputs": [
1064 |         {
1065 |           "output_type": "execute_result",
1066 |           "data": {
1067 |             "text/html": [
1068 |               "<div>\n",
1069 |               "<style scoped>\n",
1070 |               "    .dataframe tbody tr th:only-of-type {\n",
1071 |               "        vertical-align: middle;\n",
1072 |               "    }\n",
1073 |               "\n",
1074 |               "    .dataframe tbody tr th {\n",
1075 |               "        vertical-align: top;\n",
1076 |               "    }\n",
1077 |               "\n",
1078 |               "    .dataframe thead th {\n",
1079 |               "        text-align: right;\n",
1080 |               "    }\n",
1081 |               "</style>\n",
1082 |               "<table border=\"1\" class=\"dataframe\">\n",
1083 |               "  <thead>\n",
1084 |               "    <tr style=\"text-align: right;\">\n",
1085 |               "      <th></th>\n",
1086 |               "      <th>label_num</th>\n",
1087 |               "      <th>message_len</th>\n",
1088 |               "    </tr>\n",
1089 |               "  </thead>\n",
1090 |               "  <tbody>\n",
1091 |               "    <tr>\n",
1092 |               "      <th>count</th>\n",
1093 |               "      <td>4825.0</td>\n",
1094 |               "      <td>4825.000000</td>\n",
1095 |               "    </tr>\n",
1096 |               "    <tr>\n",
1097 |               "      <th>mean</th>\n",
1098 |               "      <td>0.0</td>\n",
1099 |               "      <td>71.023627</td>\n",
1100 |               "    </tr>\n",
1101 |               "    <tr>\n",
1102 |               "      <th>std</th>\n",
1103 |               "      <td>0.0</td>\n",
1104 |               "      <td>58.016023</td>\n",
1105 |               "    </tr>\n",
1106 |               "    <tr>\n",
1107 |               "      <th>min</th>\n",
1108 |               "      <td>0.0</td>\n",
1109 |               "      <td>2.000000</td>\n",
1110 |               "    </tr>\n",
1111 |               "    <tr>\n",
1112 |               "      <th>25%</th>\n",
1113 |               "      <td>0.0</td>\n",
1114 |               "      <td>33.000000</td>\n",
1115 |               "    </tr>\n",
1116 |               "    <tr>\n",
1117 |               "      <th>50%</th>\n",
1118 |               "      <td>0.0</td>\n",
1119 |               "      <td>52.000000</td>\n",
1120 |               "    </tr>\n",
1121 |               "    <tr>\n",
1122 |               "      <th>75%</th>\n",
1123 |               "      <td>0.0</td>\n",
1124 |               "      <td>92.000000</td>\n",
1125 |               "    </tr>\n",
1126 |               "    <tr>\n",
1127 |               "      <th>max</th>\n",
1128 |               "      <td>0.0</td>\n",
1129 |               "      <td>910.000000</td>\n",
1130 |               "    </tr>\n",
1131 |               "  </tbody>\n",
1132 |               "</table>\n",
1133 |               "</div>"
1134 |             ],
1135 |             "text/plain": [
1136 |               "       label_num  message_len\n",
1137 |               "count     4825.0  4825.000000\n",
1138 |               "mean         0.0    71.023627\n",
1139 |               "std          0.0    58.016023\n",
1140 |               "min          0.0     2.000000\n",
1141 |               "25%          0.0    33.000000\n",
1142 |               "50%          0.0    52.000000\n",
1143 |               "75%          0.0    92.000000\n",
1144 |               "max          0.0   910.000000"
1145 |             ]
1146 |           },
1147 |           "metadata": {
1148 |             "tags": []
1149 |           },
1150 |           "execution_count": 21
1151 |         }
1152 |       ]
1153 |     },
1154 |     {
1155 |       "cell_type": "code",
1156 |       "metadata": {
1157 |         "colab": {
1158 |           "base_uri": "https://localhost:8080/",
1159 |           "height": 286
1160 |         },
1161 |         "id": "N8qmqZyQlKcP",
1162 |         "outputId": "e11b5776-6550-414f-b815-e1b220b68408"
1163 |       },
1164 |       "source": [
1165 |         "sms[sms.label=='spam'].describe()"
1166 |       ],
1167 |       "execution_count": 22,
1168 |       "outputs": [
1169 |         {
1170 |           "output_type": "execute_result",
1171 |           "data": {
1172 |             "text/html": [
1173 |               "<div>\n",
1174 |               "<style scoped>\n",
1175 |               "    .dataframe tbody tr th:only-of-type {\n",
1176 |               "        vertical-align: middle;\n",
1177 |               "    }\n",
1178 |               "\n",
1179 |               "    .dataframe tbody tr th {\n",
1180 |               "        vertical-align: top;\n",
1181 |               "    }\n",
1182 |               "\n",
1183 |               "    .dataframe thead th {\n",
1184 |               "        text-align: right;\n",
1185 |               "    }\n",
1186 |               "</style>\n",
1187 |               "<table border=\"1\" class=\"dataframe\">\n",
1188 |               "  <thead>\n",
1189 |               "    <tr style=\"text-align: right;\">\n",
1190 |               "      <th></th>\n",
1191 |               "      <th>label_num</th>\n",
1192 |               "      <th>message_len</th>\n",
1193 |               "    </tr>\n",
1194 |               "  </thead>\n",
1195 |               "  <tbody>\n",
1196 |               "    <tr>\n",
1197 |               "      <th>count</th>\n",
1198 |               "      <td>747.0</td>\n",
1199 |               "      <td>747.000000</td>\n",
1200 |               "    </tr>\n",
1201 |               "    <tr>\n",
1202 |               "      <th>mean</th>\n",
1203 |               "      <td>1.0</td>\n",
1204 |               "      <td>138.866131</td>\n",
1205 |               "    </tr>\n",
1206 |               "    <tr>\n",
1207 |               "      <th>std</th>\n",
1208 |               "      <td>0.0</td>\n",
1209 |               "      <td>29.183082</td>\n",
1210 |               "    </tr>\n",
1211 |               "    <tr>\n",
1212 |               "      <th>min</th>\n",
1213 |               "      <td>1.0</td>\n",
1214 |               "      <td>13.000000</td>\n",
1215 |               "    </tr>\n",
1216 |               "    <tr>\n",
1217 |               "      <th>25%</th>\n",
1218 |               "      <td>1.0</td>\n",
1219 |               "      <td>132.500000</td>\n",
1220 |               "    </tr>\n",
1221 |               "    <tr>\n",
1222 |               "      <th>50%</th>\n",
1223 |               "      <td>1.0</td>\n",
1224 |               "      <td>149.000000</td>\n",
1225 |               "    </tr>\n",
1226 |               "    <tr>\n",
1227 |               "      <th>75%</th>\n",
1228 |               "      <td>1.0</td>\n",
1229 |               "      <td>157.000000</td>\n",
1230 |               "    </tr>\n",
1231 |               "    <tr>\n",
1232 |               "      <th>max</th>\n",
1233 |               "      <td>1.0</td>\n",
1234 |               "      <td>224.000000</td>\n",
1235 |               "    </tr>\n",
1236 |               "  </tbody>\n",
1237 |               "</table>\n",
1238 |               "</div>"
1239 |             ],
1240 |             "text/plain": [
1241 |               "       label_num  message_len\n",
1242 |               "count      747.0   747.000000\n",
1243 |               "mean         1.0   138.866131\n",
1244 |               "std          0.0    29.183082\n",
1245 |               "min          1.0    13.000000\n",
1246 |               "25%          1.0   132.500000\n",
1247 |               "50%          1.0   149.000000\n",
1248 |               "75%          1.0   157.000000\n",
1249 |               "max          1.0   224.000000"
1250 |             ]
1251 |           },
1252 |           "metadata": {
1253 |             "tags": []
1254 |           },
1255 |           "execution_count": 22
1256 |         }
1257 |       ]
1258 |     },
1259 |     {
1260 |       "cell_type": "code",
1261 |       "metadata": {
1262 |         "colab": {
1263 |           "base_uri": "https://localhost:8080/",
1264 |           "height": 118
1265 |         },
1266 |         "id": "kUcvNzFqlNEo",
1267 |         "outputId": "79a639f0-b7a2-4282-f31f-504f8122b851"
1268 |       },
1269 |       "source": [
1270 |         "sms[sms.message_len == 910].message.iloc[0]"
1271 |       ],
1272 |       "execution_count": 23,
1273 |       "outputs": [
1274 |         {
1275 |           "output_type": "execute_result",
1276 |           "data": {
1277 |             "application/vnd.google.colaboratory.intrinsic+json": {
1278 |               "type": "string"
1279 |             },
1280 |             "text/plain": [
1281 |               "\"For me the love should start with attraction.i should feel that I need her every time around me.she should be the first thing which comes in my thoughts.I would start the day and end it with her.she should be there every time I dream.love will be then when my every breath has her name.my life should happen around her.my life will be named to her.I would cry for her.will give all my happiness and take all her sorrows.I will be ready to fight with anyone for her.I will be in love when I will be doing the craziest things for her.love will be when I don't have to proove anyone that my girl is the most beautiful lady on the whole planet.I will always be singing praises for her.love will be when I start up making chicken curry and end up makiing sambar.life will be the most beautiful then.will get every morning and thank god for the day because she is with me.I would like to say a lot..will tell later..\""
1282 |             ]
1283 |           },
1284 |           "metadata": {
1285 |             "tags": []
1286 |           },
1287 |           "execution_count": 23
1288 |         }
1289 |       ]
1290 |     },
1291 |     {
1292 |       "cell_type": "markdown",
1293 |       "metadata": {
1294 |         "id": "7a_J_HjxlV8W"
1295 |       },
1296 |       "source": [
1297 |         "## 4. Text Pre-processing"
1298 |       ]
1299 |     },
1300 |     {
1301 |       "cell_type": "code",
1302 |       "metadata": {
1303 |         "colab": {
1304 |           "base_uri": "https://localhost:8080/",
1305 |           "height": 196
1306 |         },
1307 |         "id": "ABcHa1mslPuH",
1308 |         "outputId": "d61dd583-dedb-4fd5-e7a8-23280ad46bb7"
1309 |       },
1310 |       "source": [
1311 |         "import string\n",
1312 |         "from nltk.corpus import stopwords\n",
1313 |         "\n",
1314 |         "def text_process(mess):\n",
1315 |         "    \"\"\"\n",
1316 |         "    Takes in a string of text, then performs the following:\n",
1317 |         "    1. Remove all punctuation\n",
1318 |         "    2. Remove all stopwords\n",
1319 |         "    3. Returns a list of the cleaned text\n",
1320 |         "    \"\"\"\n",
1321 |         "    STOPWORDS = stopwords.words('english') + ['u', 'ü', 'ur', '4', '2', 'im', 'dont', 'doin', 'ure']\n",
1322 |         "    # Check characters to see if they are in punctuation\n",
1323 |         "    nopunc = [char for char in mess if char not in string.punctuation]\n",
1324 |         "\n",
1325 |         "    # Join the characters again to form the string.\n",
1326 |         "    nopunc = ''.join(nopunc)\n",
1327 |         "    \n",
1328 |         "    # Now just remove any stopwords\n",
1329 |         "    return ' '.join([word for word in nopunc.split() if word.lower() not in STOPWORDS])\n",
1330 |         "    \n",
1331 |         "sms.head()"
1332 |       ],
1333 |       "execution_count": 24,
1334 |       "outputs": [
1335 |         {
1336 |           "output_type": "execute_result",
1337 |           "data": {
1338 |             "text/html": [
1339 |               "<div>\n",
1340 |               "<style scoped>\n",
1341 |               "    .dataframe tbody tr th:only-of-type {\n",
1342 |               "        vertical-align: middle;\n",
1343 |               "    }\n",
1344 |               "\n",
1345 |               "    .dataframe tbody tr th {\n",
1346 |               "        vertical-align: top;\n",
1347 |               "    }\n",
1348 |               "\n",
1349 |               "    .dataframe thead th {\n",
1350 |               "        text-align: right;\n",
1351 |               "    }\n",
1352 |               "</style>\n",
1353 |               "<table border=\"1\" class=\"dataframe\">\n",
1354 |               "  <thead>\n",
1355 |               "    <tr style=\"text-align: right;\">\n",
1356 |               "      <th></th>\n",
1357 |               "      <th>label</th>\n",
1358 |               "      <th>message</th>\n",
1359 |               "      <th>label_num</th>\n",
1360 |               "      <th>message_len</th>\n",
1361 |               "    </tr>\n",
1362 |               "  </thead>\n",
1363 |               "  <tbody>\n",
1364 |               "    <tr>\n",
1365 |               "      <th>0</th>\n",
1366 |               "      <td>ham</td>\n",
1367 |               "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
1368 |               "      <td>0</td>\n",
1369 |               "      <td>111</td>\n",
1370 |               "    </tr>\n",
1371 |               "    <tr>\n",
1372 |               "      <th>1</th>\n",
1373 |               "      <td>ham</td>\n",
1374 |               "      <td>Ok lar... Joking wif u oni...</td>\n",
1375 |               "      <td>0</td>\n",
1376 |               "      <td>29</td>\n",
1377 |               "    </tr>\n",
1378 |               "    <tr>\n",
1379 |               "      <th>2</th>\n",
1380 |               "      <td>spam</td>\n",
1381 |               "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
1382 |               "      <td>1</td>\n",
1383 |               "      <td>155</td>\n",
1384 |               "    </tr>\n",
1385 |               "    <tr>\n",
1386 |               "      <th>3</th>\n",
1387 |               "      <td>ham</td>\n",
1388 |               "      <td>U dun say so early hor... U c already then say...</td>\n",
1389 |               "      <td>0</td>\n",
1390 |               "      <td>49</td>\n",
1391 |               "    </tr>\n",
1392 |               "    <tr>\n",
1393 |               "      <th>4</th>\n",
1394 |               "      <td>ham</td>\n",
1395 |               "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
1396 |               "      <td>0</td>\n",
1397 |               "      <td>61</td>\n",
1398 |               "    </tr>\n",
1399 |               "  </tbody>\n",
1400 |               "</table>\n",
1401 |               "</div>"
1402 |             ],
1403 |             "text/plain": [
1404 |               "  label  ... message_len\n",
1405 |               "0   ham  ...         111\n",
1406 |               "1   ham  ...          29\n",
1407 |               "2  spam  ...         155\n",
1408 |               "3   ham  ...          49\n",
1409 |               "4   ham  ...          61\n",
1410 |               "\n",
1411 |               "[5 rows x 4 columns]"
1412 |             ]
1413 |           },
1414 |           "metadata": {
1415 |             "tags": []
1416 |           },
1417 |           "execution_count": 24
1418 |         }
1419 |       ]
1420 |     },
1421 |     {
1422 |       "cell_type": "code",
1423 |       "metadata": {
1424 |         "colab": {
1425 |           "base_uri": "https://localhost:8080/",
1426 |           "height": 196
1427 |         },
1428 |         "id": "bujuGVqSlcFn",
1429 |         "outputId": "3763a794-c1b9-4262-c97a-5dd7656a7330"
1430 |       },
1431 |       "source": [
1432 |         "sms['clean_msg'] = sms.message.apply(text_process)\n",
1433 |         "sms.head()"
1434 |       ],
1435 |       "execution_count": 28,
1436 |       "outputs": [
1437 |         {
1438 |           "output_type": "execute_result",
1439 |           "data": {
1440 |             "text/html": [
1441 |               "<div>\n",
1442 |               "<style scoped>\n",
1443 |               "    .dataframe tbody tr th:only-of-type {\n",
1444 |               "        vertical-align: middle;\n",
1445 |               "    }\n",
1446 |               "\n",
1447 |               "    .dataframe tbody tr th {\n",
1448 |               "        vertical-align: top;\n",
1449 |               "    }\n",
1450 |               "\n",
1451 |               "    .dataframe thead th {\n",
1452 |               "        text-align: right;\n",
1453 |               "    }\n",
1454 |               "</style>\n",
1455 |               "<table border=\"1\" class=\"dataframe\">\n",
1456 |               "  <thead>\n",
1457 |               "    <tr style=\"text-align: right;\">\n",
1458 |               "      <th></th>\n",
1459 |               "      <th>label</th>\n",
1460 |               "      <th>message</th>\n",
1461 |               "      <th>label_num</th>\n",
1462 |               "      <th>message_len</th>\n",
1463 |               "      <th>clean_msg</th>\n",
1464 |               "    </tr>\n",
1465 |               "  </thead>\n",
1466 |               "  <tbody>\n",
1467 |               "    <tr>\n",
1468 |               "      <th>0</th>\n",
1469 |               "      <td>ham</td>\n",
1470 |               "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
1471 |               "      <td>0</td>\n",
1472 |               "      <td>111</td>\n",
1473 |               "      <td>Go jurong point crazy Available bugis n great ...</td>\n",
1474 |               "    </tr>\n",
1475 |               "    <tr>\n",
1476 |               "      <th>1</th>\n",
1477 |               "      <td>ham</td>\n",
1478 |               "      <td>Ok lar... Joking wif u oni...</td>\n",
1479 |               "      <td>0</td>\n",
1480 |               "      <td>29</td>\n",
1481 |               "      <td>Ok lar Joking wif oni</td>\n",
1482 |               "    </tr>\n",
1483 |               "    <tr>\n",
1484 |               "      <th>2</th>\n",
1485 |               "      <td>spam</td>\n",
1486 |               "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
1487 |               "      <td>1</td>\n",
1488 |               "      <td>155</td>\n",
1489 |               "      <td>Free entry wkly comp win FA Cup final tkts 21s...</td>\n",
1490 |               "    </tr>\n",
1491 |               "    <tr>\n",
1492 |               "      <th>3</th>\n",
1493 |               "      <td>ham</td>\n",
1494 |               "      <td>U dun say so early hor... U c already then say...</td>\n",
1495 |               "      <td>0</td>\n",
1496 |               "      <td>49</td>\n",
1497 |               "      <td>dun say early hor c already say</td>\n",
1498 |               "    </tr>\n",
1499 |               "    <tr>\n",
1500 |               "      <th>4</th>\n",
1501 |               "      <td>ham</td>\n",
1502 |               "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
1503 |               "      <td>0</td>\n",
1504 |               "      <td>61</td>\n",
1505 |               "      <td>Nah think goes usf lives around though</td>\n",
1506 |               "    </tr>\n",
1507 |               "  </tbody>\n",
1508 |               "</table>\n",
1509 |               "</div>"
1510 |             ],
1511 |             "text/plain": [
1512 |               "  label  ...                                          clean_msg\n",
1513 |               "0   ham  ...  Go jurong point crazy Available bugis n great ...\n",
1514 |               "1   ham  ...                              Ok lar Joking wif oni\n",
1515 |               "2  spam  ...  Free entry wkly comp win FA Cup final tkts 21s...\n",
1516 |               "3   ham  ...                    dun say early hor c already say\n",
1517 |               "4   ham  ...             Nah think goes usf lives around though\n",
1518 |               "\n",
1519 |               "[5 rows x 5 columns]"
1520 |             ]
1521 |           },
1522 |           "metadata": {
1523 |             "tags": []
1524 |           },
1525 |           "execution_count": 28
1526 |         }
1527 |       ]
1528 |     },
1529 |     {
1530 |       "cell_type": "code",
1531 |       "metadata": {
1532 |         "colab": {
1533 |           "base_uri": "https://localhost:8080/"
1534 |         },
1535 |         "id": "r_n-DKwelfan",
1536 |         "outputId": "3feee23f-e322-4cb1-82f5-273229cedaa1"
1537 |       },
1538 |       "source": [
1539 |         "type(stopwords.words('english'))"
1540 |       ],
1541 |       "execution_count": 29,
1542 |       "outputs": [
1543 |         {
1544 |           "output_type": "execute_result",
1545 |           "data": {
1546 |             "text/plain": [
1547 |               "list"
1548 |             ]
1549 |           },
1550 |           "metadata": {
1551 |             "tags": []
1552 |           },
1553 |           "execution_count": 29
1554 |         }
1555 |       ]
1556 |     },
1557 |     {
1558 |       "cell_type": "code",
1559 |       "metadata": {
1560 |         "colab": {
1561 |           "base_uri": "https://localhost:8080/"
1562 |         },
1563 |         "id": "yNrGdgljlsAx",
1564 |         "outputId": "38f42f2e-17ae-4ee4-e4c2-43ac5a10d4cb"
1565 |       },
1566 |       "source": [
1567 |         "from collections import Counter\n",
1568 |         "\n",
1569 |         "words = sms[sms.label=='ham'].clean_msg.apply(lambda x: [word.lower() for word in x.split()])\n",
1570 |         "ham_words = Counter()\n",
1571 |         "\n",
1572 |         "for msg in words:\n",
1573 |         "    ham_words.update(msg)\n",
1574 |         "    \n",
1575 |         "print(ham_words.most_common(50))"
1576 |       ],
1577 |       "execution_count": 30,
1578 |       "outputs": [
1579 |         {
1580 |           "output_type": "stream",
1581 |           "text": [
1582 |             "[('get', 303), ('ltgt', 276), ('ok', 272), ('go', 247), ('ill', 236), ('know', 232), ('got', 231), ('like', 229), ('call', 229), ('come', 224), ('good', 222), ('time', 189), ('day', 187), ('love', 185), ('going', 167), ('want', 163), ('one', 162), ('home', 160), ('lor', 160), ('need', 156), ('sorry', 153), ('still', 146), ('see', 137), ('n', 134), ('later', 134), ('da', 131), ('r', 131), ('back', 129), ('think', 128), ('well', 126), ('today', 125), ('send', 123), ('tell', 121), ('cant', 118), ('ì', 117), ('hi', 117), ('take', 112), ('much', 112), ('oh', 111), ('night', 107), ('hey', 106), ('happy', 105), ('great', 100), ('way', 100), ('hope', 99), ('pls', 98), ('work', 96), ('wat', 95), ('thats', 94), ('dear', 94)]\n"
1583 |           ],
1584 |           "name": "stdout"
1585 |         }
1586 |       ]
1587 |     },
1588 |     {
1589 |       "cell_type": "code",
1590 |       "metadata": {
1591 |         "colab": {
1592 |           "base_uri": "https://localhost:8080/"
1593 |         },
1594 |         "id": "3KVBmlrIlujl",
1595 |         "outputId": "fb138d9a-51fc-458a-99b5-74d6d81629d6"
1596 |       },
1597 |       "source": [
1598 |         "words = sms[sms.label=='spam'].clean_msg.apply(lambda x: [word.lower() for word in x.split()])\n",
1599 |         "spam_words = Counter()\n",
1600 |         "\n",
1601 |         "for msg in words:\n",
1602 |         "    spam_words.update(msg)\n",
1603 |         "    \n",
1604 |         "print(spam_words.most_common(50))"
1605 |       ],
1606 |       "execution_count": 31,
1607 |       "outputs": [
1608 |         {
1609 |           "output_type": "stream",
1610 |           "text": [
1611 |             "[('call', 347), ('free', 216), ('txt', 150), ('mobile', 123), ('text', 120), ('claim', 113), ('stop', 113), ('reply', 101), ('prize', 92), ('get', 83), ('new', 69), ('send', 67), ('nokia', 65), ('urgent', 63), ('cash', 62), ('win', 60), ('contact', 56), ('service', 55), ('please', 52), ('guaranteed', 50), ('customer', 49), ('16', 49), ('week', 49), ('tone', 48), ('per', 46), ('phone', 45), ('18', 43), ('chat', 42), ('awarded', 38), ('draw', 38), ('latest', 36), ('å£1000', 35), ('line', 35), ('150ppm', 34), ('mins', 34), ('receive', 33), ('camera', 33), ('1', 33), ('every', 33), ('message', 32), ('holiday', 32), ('landline', 32), ('shows', 31), ('å£2000', 31), ('go', 31), ('box', 30), ('number', 30), ('apply', 29), ('code', 29), ('live', 29)]\n"
1612 |           ],
1613 |           "name": "stdout"
1614 |         }
1615 |       ]
1616 |     },
1617 |     {
1618 |       "cell_type": "code",
1619 |       "metadata": {
1620 |         "colab": {
1621 |           "base_uri": "https://localhost:8080/"
1622 |         },
1623 |         "id": "IeKB4sQtlwxj",
1624 |         "outputId": "553294bd-30e7-4e8d-cacc-185a42665a8d"
1625 |       },
1626 |       "source": [
1627 |         "# how to define X and y (from the SMS data) for use with COUNTVECTORIZER\n",
1628 |         "X = sms.clean_msg\n",
1629 |         "y = sms.label_num\n",
1630 |         "print(X.shape)\n",
1631 |         "print(y.shape)"
1632 |       ],
1633 |       "execution_count": 32,
1634 |       "outputs": [
1635 |         {
1636 |           "output_type": "stream",
1637 |           "text": [
1638 |             "(5572,)\n",
1639 |             "(5572,)\n"
1640 |           ],
1641 |           "name": "stdout"
1642 |         }
1643 |       ]
1644 |     },
1645 |     {
1646 |       "cell_type": "code",
1647 |       "metadata": {
1648 |         "colab": {
1649 |           "base_uri": "https://localhost:8080/"
1650 |         },
1651 |         "id": "qwXh2kDel0a3",
1652 |         "outputId": "c6395419-8fac-47cd-c02b-a97b964d63e1"
1653 |       },
1654 |       "source": [
1655 |         "# split X and y into training and testing sets \n",
1656 |         "from sklearn.model_selection import train_test_split\n",
1657 |         "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=1)\n",
1658 |         "print(X_train.shape)\n",
1659 |         "print(X_test.shape)\n",
1660 |         "print(y_train.shape)\n",
1661 |         "print(y_test.shape)"
1662 |       ],
1663 |       "execution_count": 33,
1664 |       "outputs": [
1665 |         {
1666 |           "output_type": "stream",
1667 |           "text": [
1668 |             "(4179,)\n",
1669 |             "(1393,)\n",
1670 |             "(4179,)\n",
1671 |             "(1393,)\n"
1672 |           ],
1673 |           "name": "stdout"
1674 |         }
1675 |       ]
1676 |     },
1677 |     {
1678 |       "cell_type": "code",
1679 |       "metadata": {
1680 |         "colab": {
1681 |           "base_uri": "https://localhost:8080/"
1682 |         },
1683 |         "id": "9CdsKNV9l6Ij",
1684 |         "outputId": "554084da-ec1b-4311-caee-e1ef4faea2dd"
1685 |       },
1686 |       "source": [
1687 |         "from sklearn.feature_extraction.text import CountVectorizer\n",
1688 |         "\n",
1689 |         "# instantiate the vectorizer\n",
1690 |         "vect = CountVectorizer()\n",
1691 |         "vect.fit(X_train)"
1692 |       ],
1693 |       "execution_count": 34,
1694 |       "outputs": [
1695 |         {
1696 |           "output_type": "execute_result",
1697 |           "data": {
1698 |             "text/plain": [
1699 |               "CountVectorizer(analyzer='word', binary=False, decode_error='strict',\n",
1700 |               "                dtype=<class 'numpy.int64'>, encoding='utf-8', input='content',\n",
1701 |               "                lowercase=True, max_df=1.0, max_features=None, min_df=1,\n",
1702 |               "                ngram_range=(1, 1), preprocessor=None, stop_words=None,\n",
1703 |               "                strip_accents=None, token_pattern='(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
1704 |               "                tokenizer=None, vocabulary=None)"
1705 |             ]
1706 |           },
1707 |           "metadata": {
1708 |             "tags": []
1709 |           },
1710 |           "execution_count": 34
1711 |         }
1712 |       ]
1713 |     },
1714 |     {
1715 |       "cell_type": "code",
1716 |       "metadata": {
1717 |         "id": "RfkRg3e-l8Wy"
1718 |       },
1719 |       "source": [
1720 |         "# learn training data vocabulary, then use it to create a document-term matrix\n",
1721 |         "X_train_dtm = vect.transform(X_train)"
1722 |       ],
1723 |       "execution_count": 35,
1724 |       "outputs": []
1725 |     },
1726 |     {
1727 |       "cell_type": "code",
1728 |       "metadata": {
1729 |         "id": "1AoyWLyhl_aR"
1730 |       },
1731 |       "source": [
1732 |         "# equivalently: combine fit and transform into a single step\n",
1733 |         "X_train_dtm = vect.fit_transform(X_train)"
1734 |       ],
1735 |       "execution_count": 36,
1736 |       "outputs": []
1737 |     },
1738 |     {
1739 |       "cell_type": "code",
1740 |       "metadata": {
1741 |         "colab": {
1742 |           "base_uri": "https://localhost:8080/"
1743 |         },
1744 |         "id": "dnx4f9FrmCyI",
1745 |         "outputId": "fe20bf55-55ca-4278-c068-0f9d3ea64927"
1746 |       },
1747 |       "source": [
1748 |         "# examine the document-term matrix\n",
1749 |         "X_train_dtm"
1750 |       ],
1751 |       "execution_count": 37,
1752 |       "outputs": [
1753 |         {
1754 |           "output_type": "execute_result",
1755 |           "data": {
1756 |             "text/plain": [
1757 |               "<4179x7996 sparse matrix of type '<class 'numpy.int64'>'\n",
1758 |               "\twith 34796 stored elements in Compressed Sparse Row format>"
1759 |             ]
1760 |           },
1761 |           "metadata": {
1762 |             "tags": []
1763 |           },
1764 |           "execution_count": 37
1765 |         }
1766 |       ]
1767 |     },
1768 |     {
1769 |       "cell_type": "code",
1770 |       "metadata": {
1771 |         "colab": {
1772 |           "base_uri": "https://localhost:8080/"
1773 |         },
1774 |         "id": "v5Tc7njsmE-7",
1775 |         "outputId": "e2fa087f-06f9-494f-d4ad-3f9662fdc624"
1776 |       },
1777 |       "source": [
1778 |         "# transform testing data (using fitted vocabulary) into a document-term matrix\n",
1779 |         "X_test_dtm = vect.transform(X_test)\n",
1780 |         "X_test_dtm"
1781 |       ],
1782 |       "execution_count": 38,
1783 |       "outputs": [
1784 |         {
1785 |           "output_type": "execute_result",
1786 |           "data": {
1787 |             "text/plain": [
1788 |               "<1393x7996 sparse matrix of type '<class 'numpy.int64'>'\n",
1789 |               "\twith 9971 stored elements in Compressed Sparse Row format>"
1790 |             ]
1791 |           },
1792 |           "metadata": {
1793 |             "tags": []
1794 |           },
1795 |           "execution_count": 38
1796 |         }
1797 |       ]
1798 |     },
1799 |     {
1800 |       "cell_type": "code",
1801 |       "metadata": {
1802 |         "colab": {
1803 |           "base_uri": "https://localhost:8080/"
1804 |         },
1805 |         "id": "KLXHpl3cmHSF",
1806 |         "outputId": "bd78b3f7-d3c8-45f4-e056-525148d962bb"
1807 |       },
1808 |       "source": [
1809 |         "from sklearn.feature_extraction.text import TfidfTransformer\n",
1810 |         "\n",
1811 |         "tfidf_transformer = TfidfTransformer()\n",
1812 |         "tfidf_transformer.fit(X_train_dtm)\n",
1813 |         "tfidf_transformer.transform(X_train_dtm)"
1814 |       ],
1815 |       "execution_count": 39,
1816 |       "outputs": [
1817 |         {
1818 |           "output_type": "execute_result",
1819 |           "data": {
1820 |             "text/plain": [
1821 |               "<4179x7996 sparse matrix of type '<class 'numpy.float64'>'\n",
1822 |               "\twith 34796 stored elements in Compressed Sparse Row format>"
1823 |             ]
1824 |           },
1825 |           "metadata": {
1826 |             "tags": []
1827 |           },
1828 |           "execution_count": 39
1829 |         }
1830 |       ]
1831 |     },
1832 |     {
1833 |       "cell_type": "markdown",
1834 |       "metadata": {
1835 |         "id": "5jrMB2sGmNtH"
1836 |       },
1837 |       "source": [
1838 |         "## 5. Building and evaluating a model"
1839 |       ]
1840 |     },
1841 |     {
1842 |       "cell_type": "code",
1843 |       "metadata": {
1844 |         "colab": {
1845 |           "base_uri": "https://localhost:8080/"
1846 |         },
1847 |         "id": "_UA_grwfmKAx",
1848 |         "outputId": "68f9c903-b1bf-4109-c056-d7e7e70e44cd"
1849 |       },
1850 |       "source": [
1851 |         "# import and instantiate a Multinomial Naive Bayes model\n",
1852 |         "from sklearn.naive_bayes import MultinomialNB\n",
1853 |         "nb = MultinomialNB()\n",
1854 |         "# train the model using X_train_dtm (timing it with an IPython \"magic command\")\n",
1855 |         "%time nb.fit(X_train_dtm, y_train)"
1856 |       ],
1857 |       "execution_count": 40,
1858 |       "outputs": [
1859 |         {
1860 |           "output_type": "stream",
1861 |           "text": [
1862 |             "CPU times: user 6.55 ms, sys: 0 ns, total: 6.55 ms\n",
1863 |             "Wall time: 9.38 ms\n"
1864 |           ],
1865 |           "name": "stdout"
1866 |         },
1867 |         {
1868 |           "output_type": "execute_result",
1869 |           "data": {
1870 |             "text/plain": [
1871 |               "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)"
1872 |             ]
1873 |           },
1874 |           "metadata": {
1875 |             "tags": []
1876 |           },
1877 |           "execution_count": 40
1878 |         }
1879 |       ]
1880 |     },
1881 |     {
1882 |       "cell_type": "code",
1883 |       "metadata": {
1884 |         "colab": {
1885 |           "base_uri": "https://localhost:8080/"
1886 |         },
1887 |         "id": "sX3Xtcn8mQ6F",
1888 |         "outputId": "07e0995c-2596-41d3-b467-a1039e6300cc"
1889 |       },
1890 |       "source": [
1891 |         "# make class predictions for X_test_dtm\n",
1892 |         "y_pred_class = nb.predict(X_test_dtm)\n",
1893 |         "# calculate accuracy of class predictions\n",
1894 |         "from sklearn import metrics\n",
1895 |         "metrics.accuracy_score(y_test, y_pred_class)"
1896 |       ],
1897 |       "execution_count": 41,
1898 |       "outputs": [
1899 |         {
1900 |           "output_type": "execute_result",
1901 |           "data": {
1902 |             "text/plain": [
1903 |               "0.9827709978463748"
1904 |             ]
1905 |           },
1906 |           "metadata": {
1907 |             "tags": []
1908 |           },
1909 |           "execution_count": 41
1910 |         }
1911 |       ]
1912 |     },
1913 |     {
1914 |       "cell_type": "code",
1915 |       "metadata": {
1916 |         "colab": {
1917 |           "base_uri": "https://localhost:8080/"
1918 |         },
1919 |         "id": "Zs88HASKmUOy",
1920 |         "outputId": "500deb05-b52e-4c8c-b171-d01de5050da4"
1921 |       },
1922 |       "source": [
1923 |         "# print the confusion matrix\n",
1924 |         "metrics.confusion_matrix(y_test, y_pred_class)"
1925 |       ],
1926 |       "execution_count": 42,
1927 |       "outputs": [
1928 |         {
1929 |           "output_type": "execute_result",
1930 |           "data": {
1931 |             "text/plain": [
1932 |               "array([[1205,    8],\n",
1933 |               "       [  16,  164]])"
1934 |             ]
1935 |           },
1936 |           "metadata": {
1937 |             "tags": []
1938 |           },
1939 |           "execution_count": 42
1940 |         }
1941 |       ]
1942 |     },
1943 |     {
1944 |       "cell_type": "code",
1945 |       "metadata": {
1946 |         "colab": {
1947 |           "base_uri": "https://localhost:8080/"
1948 |         },
1949 |         "id": "o_UjIBNEmW1Q",
1950 |         "outputId": "899110aa-c49c-463c-df46-cd03830198f7"
1951 |       },
1952 |       "source": [
1953 |         "X_test.shape"
1954 |       ],
1955 |       "execution_count": 43,
1956 |       "outputs": [
1957 |         {
1958 |           "output_type": "execute_result",
1959 |           "data": {
1960 |             "text/plain": [
1961 |               "(1393,)"
1962 |             ]
1963 |           },
1964 |           "metadata": {
1965 |             "tags": []
1966 |           },
1967 |           "execution_count": 43
1968 |         }
1969 |       ]
1970 |     },
1971 |     {
1972 |       "cell_type": "code",
1973 |       "metadata": {
1974 |         "colab": {
1975 |           "base_uri": "https://localhost:8080/"
1976 |         },
1977 |         "id": "d28YU2iUmZXm",
1978 |         "outputId": "47bb2ae0-2e40-44e5-95ed-8aba3975f965"
1979 |       },
1980 |       "source": [
1981 |         "X_test[y_pred_class > y_test]"
1982 |       ],
1983 |       "execution_count": 44,
1984 |       "outputs": [
1985 |         {
1986 |           "output_type": "execute_result",
1987 |           "data": {
1988 |             "text/plain": [
1989 |               "2418    Madamregret disturbancemight receive reference...\n",
1990 |               "4598                                laid airtel line rest\n",
1991 |               "386                                   Customer place call\n",
1992 |               "1289    HeyGreat dealFarm tour 9am 5pm 95pax 50 deposi...\n",
1993 |               "5094    Hi ShanilRakhesh herethanksi exchanged uncut d...\n",
1994 |               "494                                      free nowcan call\n",
1995 |               "759     Call youcarlos isare phones vibrate acting mig...\n",
1996 |               "3140                                  Customer place call\n",
1997 |               "Name: clean_msg, dtype: object"
1998 |             ]
1999 |           },
2000 |           "metadata": {
2001 |             "tags": []
2002 |           },
2003 |           "execution_count": 44
2004 |         }
2005 |       ]
2006 |     },
2007 |     {
2008 |       "cell_type": "code",
2009 |       "metadata": {
2010 |         "colab": {
2011 |           "base_uri": "https://localhost:8080/"
2012 |         },
2013 |         "id": "QCdvZlsNmfY6",
2014 |         "outputId": "3adc09a8-6bd6-476f-da06-253ac0d5dfaa"
2015 |       },
2016 |       "source": [
2017 |         "# print message text for false negatives (spam incorrectly classifier)\n",
2018 |         "X_test[y_pred_class < y_test]"
2019 |       ],
2020 |       "execution_count": 45,
2021 |       "outputs": [
2022 |         {
2023 |           "output_type": "execute_result",
2024 |           "data": {
2025 |             "text/plain": [
2026 |               "4674    Hi babe Chloe r smashed saturday night great w...\n",
2027 |               "3528    Xmas New Years Eve tickets sale club day 10am ...\n",
2028 |               "3417    LIFE never much fun great came made truly spec...\n",
2029 |               "2773    come takes little time child afraid dark becom...\n",
2030 |               "1960    Guess Somebody know secretly fancies Wanna fin...\n",
2031 |               "5       FreeMsg Hey darling 3 weeks word back Id like ...\n",
2032 |               "2078                         85233 FREERingtoneReply REAL\n",
2033 |               "1457    CLAIRE havin borin time alone wanna cum 2nite ...\n",
2034 |               "190     unique enough Find 30th August wwwareyouunique...\n",
2035 |               "2429    Guess IThis first time created web page WWWASJ...\n",
2036 |               "3057    unsubscribed services Get tons sexy babes hunk...\n",
2037 |               "1021    Guess Somebody know secretly fancies Wanna fin...\n",
2038 |               "4067    TBSPERSOLVO chasing us since Sept forå£38 defi...\n",
2039 |               "3358         Sorry missed call lets talk time 07090201529\n",
2040 |               "2821    ROMCAPspam Everyone around responding well pre...\n",
2041 |               "2247    Back work 2morro half term C 2nite sexy passio...\n",
2042 |               "Name: clean_msg, dtype: object"
2043 |             ]
2044 |           },
2045 |           "metadata": {
2046 |             "tags": []
2047 |           },
2048 |           "execution_count": 45
2049 |         }
2050 |       ]
2051 |     },
2052 |     {
2053 |       "cell_type": "code",
2054 |       "metadata": {
2055 |         "colab": {
2056 |           "base_uri": "https://localhost:8080/",
2057 |           "height": 35
2058 |         },
2059 |         "id": "ptFhEiLimiCZ",
2060 |         "outputId": "514a66db-36a0-4214-bb3f-a774a9b82e72"
2061 |       },
2062 |       "source": [
2063 |         "# example of false negative \n",
2064 |         "X_test[4949]"
2065 |       ],
2066 |       "execution_count": 46,
2067 |       "outputs": [
2068 |         {
2069 |           "output_type": "execute_result",
2070 |           "data": {
2071 |             "application/vnd.google.colaboratory.intrinsic+json": {
2072 |               "type": "string"
2073 |             },
2074 |             "text/plain": [
2075 |               "'Hi probably much fun get message thought id txt cos bored james farting night'"
2076 |             ]
2077 |           },
2078 |           "metadata": {
2079 |             "tags": []
2080 |           },
2081 |           "execution_count": 46
2082 |         }
2083 |       ]
2084 |     },
2085 |     {
2086 |       "cell_type": "code",
2087 |       "metadata": {
2088 |         "colab": {
2089 |           "base_uri": "https://localhost:8080/"
2090 |         },
2091 |         "id": "KCIj1ItimlhQ",
2092 |         "outputId": "f7a656ee-0d05-448f-8b7b-69c90a5be155"
2093 |       },
2094 |       "source": [
2095 |         "# calculate predicted probabilities for X_test_dtm (poorly calibrated)\n",
2096 |         "y_pred_prob = nb.predict_proba(X_test_dtm)[:, 1]\n",
2097 |         "y_pred_prob"
2098 |       ],
2099 |       "execution_count": 47,
2100 |       "outputs": [
2101 |         {
2102 |           "output_type": "execute_result",
2103 |           "data": {
2104 |             "text/plain": [
2105 |               "array([2.11903975e-02, 3.97831612e-04, 1.06470895e-03, ...,\n",
2106 |               "       1.31939653e-02, 9.99821127e-05, 6.04083365e-06])"
2107 |             ]
2108 |           },
2109 |           "metadata": {
2110 |             "tags": []
2111 |           },
2112 |           "execution_count": 47
2113 |         }
2114 |       ]
2115 |     },
2116 |     {
2117 |       "cell_type": "code",
2118 |       "metadata": {
2119 |         "colab": {
2120 |           "base_uri": "https://localhost:8080/"
2121 |         },
2122 |         "id": "wZjMVAZemomm",
2123 |         "outputId": "43ec9b84-3d96-4065-e066-d4bd8eb15b5f"
2124 |       },
2125 |       "source": [
2126 |         "# calculate AUC\n",
2127 |         "metrics.roc_auc_score(y_test, y_pred_prob)"
2128 |       ],
2129 |       "execution_count": 48,
2130 |       "outputs": [
2131 |         {
2132 |           "output_type": "execute_result",
2133 |           "data": {
2134 |             "text/plain": [
2135 |               "0.9774342768159751"
2136 |             ]
2137 |           },
2138 |           "metadata": {
2139 |             "tags": []
2140 |           },
2141 |           "execution_count": 48
2142 |         }
2143 |       ]
2144 |     },
2145 |     {
2146 |       "cell_type": "code",
2147 |       "metadata": {
2148 |         "colab": {
2149 |           "base_uri": "https://localhost:8080/"
2150 |         },
2151 |         "id": "sfeq_Xr7mrDm",
2152 |         "outputId": "d607a8ec-7303-45a2-eb7e-3fa67146a39e"
2153 |       },
2154 |       "source": [
2155 |         "from sklearn.feature_extraction.text import TfidfTransformer\n",
2156 |         "from sklearn.pipeline import Pipeline\n",
2157 |         "\n",
2158 |         "pipe = Pipeline([('bow', CountVectorizer()), \n",
2159 |         "                 ('tfid', TfidfTransformer()),  \n",
2160 |         "                 ('model', MultinomialNB())])\n",
2161 |         "pipe.fit(X_train, y_train)"
2162 |       ],
2163 |       "execution_count": 49,
2164 |       "outputs": [
2165 |         {
2166 |           "output_type": "execute_result",
2167 |           "data": {
2168 |             "text/plain": [
2169 |               "Pipeline(memory=None,\n",
2170 |               "         steps=[('bow',\n",
2171 |               "                 CountVectorizer(analyzer='word', binary=False,\n",
2172 |               "                                 decode_error='strict',\n",
2173 |               "                                 dtype=<class 'numpy.int64'>, encoding='utf-8',\n",
2174 |               "                                 input='content', lowercase=True, max_df=1.0,\n",
2175 |               "                                 max_features=None, min_df=1,\n",
2176 |               "                                 ngram_range=(1, 1), preprocessor=None,\n",
2177 |               "                                 stop_words=None, strip_accents=None,\n",
2178 |               "                                 token_pattern='(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
2179 |               "                                 tokenizer=None, vocabulary=None)),\n",
2180 |               "                ('tfid',\n",
2181 |               "                 TfidfTransformer(norm='l2', smooth_idf=True,\n",
2182 |               "                                  sublinear_tf=False, use_idf=True)),\n",
2183 |               "                ('model',\n",
2184 |               "                 MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True))],\n",
2185 |               "         verbose=False)"
2186 |             ]
2187 |           },
2188 |           "metadata": {
2189 |             "tags": []
2190 |           },
2191 |           "execution_count": 49
2192 |         }
2193 |       ]
2194 |     },
2195 |     {
2196 |       "cell_type": "code",
2197 |       "metadata": {
2198 |         "colab": {
2199 |           "base_uri": "https://localhost:8080/"
2200 |         },
2201 |         "id": "VXV7G_RtmtuI",
2202 |         "outputId": "c0f9ab52-d7a3-4573-d06e-5d66d22af906"
2203 |       },
2204 |       "source": [
2205 |         "y_pred = pipe.predict(X_test)\n",
2206 |         "metrics.accuracy_score(y_test, y_pred)"
2207 |       ],
2208 |       "execution_count": 50,
2209 |       "outputs": [
2210 |         {
2211 |           "output_type": "execute_result",
2212 |           "data": {
2213 |             "text/plain": [
2214 |               "0.9669777458722182"
2215 |             ]
2216 |           },
2217 |           "metadata": {
2218 |             "tags": []
2219 |           },
2220 |           "execution_count": 50
2221 |         }
2222 |       ]
2223 |     },
2224 |     {
2225 |       "cell_type": "code",
2226 |       "metadata": {
2227 |         "colab": {
2228 |           "base_uri": "https://localhost:8080/"
2229 |         },
2230 |         "id": "gPRQSIP-mwym",
2231 |         "outputId": "5fd9f66a-3b74-43bb-e577-c1720abba3f3"
2232 |       },
2233 |       "source": [
2234 |         "metrics.confusion_matrix(y_test, y_pred)"
2235 |       ],
2236 |       "execution_count": 51,
2237 |       "outputs": [
2238 |         {
2239 |           "output_type": "execute_result",
2240 |           "data": {
2241 |             "text/plain": [
2242 |               "array([[1213,    0],\n",
2243 |               "       [  46,  134]])"
2244 |             ]
2245 |           },
2246 |           "metadata": {
2247 |             "tags": []
2248 |           },
2249 |           "execution_count": 51
2250 |         }
2251 |       ]
2252 |     },
2253 |     {
2254 |       "cell_type": "markdown",
2255 |       "metadata": {
2256 |         "id": "yZW33zDRm3Lz"
2257 |       },
2258 |       "source": [
2259 |         "## 6. Comparing models"
2260 |       ]
2261 |     },
2262 |     {
2263 |       "cell_type": "code",
2264 |       "metadata": {
2265 |         "colab": {
2266 |           "base_uri": "https://localhost:8080/"
2267 |         },
2268 |         "id": "Bgl9ffKdmz8O",
2269 |         "outputId": "8bd03dd4-97fb-45d8-ff03-d8252b6ac139"
2270 |       },
2271 |       "source": [
2272 |         "# import an instantiate a logistic regression model\n",
2273 |         "from sklearn.linear_model import LogisticRegression\n",
2274 |         "logreg = LogisticRegression(solver='liblinear')\n",
2275 |         "# train the model using X_train_dtm\n",
2276 |         "%time logreg.fit(X_train_dtm, y_train)"
2277 |       ],
2278 |       "execution_count": 52,
2279 |       "outputs": [
2280 |         {
2281 |           "output_type": "stream",
2282 |           "text": [
2283 |             "CPU times: user 19.8 ms, sys: 3.81 ms, total: 23.6 ms\n",
2284 |             "Wall time: 29 ms\n"
2285 |           ],
2286 |           "name": "stdout"
2287 |         },
2288 |         {
2289 |           "output_type": "execute_result",
2290 |           "data": {
2291 |             "text/plain": [
2292 |               "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
2293 |               "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
2294 |               "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
2295 |               "                   random_state=None, solver='liblinear', tol=0.0001, verbose=0,\n",
2296 |               "                   warm_start=False)"
2297 |             ]
2298 |           },
2299 |           "metadata": {
2300 |             "tags": []
2301 |           },
2302 |           "execution_count": 52
2303 |         }
2304 |       ]
2305 |     },
2306 |     {
2307 |       "cell_type": "code",
2308 |       "metadata": {
2309 |         "colab": {
2310 |           "base_uri": "https://localhost:8080/"
2311 |         },
2312 |         "id": "E-EYcA-sm6Uu",
2313 |         "outputId": "9b11a513-b2f4-490a-b336-048b4ce177f7"
2314 |       },
2315 |       "source": [
2316 |         "# make class predictions for X_test_dtm\n",
2317 |         "y_pred_class = logreg.predict(X_test_dtm)\n",
2318 |         "# calculate predicted probabilities for X_test_dtm (well calibrated)\n",
2319 |         "y_pred_prob = logreg.predict_proba(X_test_dtm)[:, 1]\n",
2320 |         "y_pred_prob"
2321 |       ],
2322 |       "execution_count": 53,
2323 |       "outputs": [
2324 |         {
2325 |           "output_type": "execute_result",
2326 |           "data": {
2327 |             "text/plain": [
2328 |               "array([0.01694418, 0.0152182 , 0.08261755, ..., 0.02198942, 0.00531726,\n",
2329 |               "       0.00679188])"
2330 |             ]
2331 |           },
2332 |           "metadata": {
2333 |             "tags": []
2334 |           },
2335 |           "execution_count": 53
2336 |         }
2337 |       ]
2338 |     },
2339 |     {
2340 |       "cell_type": "code",
2341 |       "metadata": {
2342 |         "colab": {
2343 |           "base_uri": "https://localhost:8080/"
2344 |         },
2345 |         "id": "gm65nFGkm9fB",
2346 |         "outputId": "0f1b8e50-1aa6-4544-9787-533e26a097ea"
2347 |       },
2348 |       "source": [
2349 |         "# calculate accuracy\n",
2350 |         "metrics.accuracy_score(y_test, y_pred_class)"
2351 |       ],
2352 |       "execution_count": 54,
2353 |       "outputs": [
2354 |         {
2355 |           "output_type": "execute_result",
2356 |           "data": {
2357 |             "text/plain": [
2358 |               "0.9842067480258435"
2359 |             ]
2360 |           },
2361 |           "metadata": {
2362 |             "tags": []
2363 |           },
2364 |           "execution_count": 54
2365 |         }
2366 |       ]
2367 |     },
2368 |     {
2369 |       "cell_type": "code",
2370 |       "metadata": {
2371 |         "colab": {
2372 |           "base_uri": "https://localhost:8080/"
2373 |         },
2374 |         "id": "UXP8643anACj",
2375 |         "outputId": "45a3e687-769a-4bb0-f3c3-f01c4af370fa"
2376 |       },
2377 |       "source": [
2378 |         "metrics.confusion_matrix(y_test, y_pred_class)"
2379 |       ],
2380 |       "execution_count": 55,
2381 |       "outputs": [
2382 |         {
2383 |           "output_type": "execute_result",
2384 |           "data": {
2385 |             "text/plain": [
2386 |               "array([[1213,    0],\n",
2387 |               "       [  22,  158]])"
2388 |             ]
2389 |           },
2390 |           "metadata": {
2391 |             "tags": []
2392 |           },
2393 |           "execution_count": 55
2394 |         }
2395 |       ]
2396 |     },
2397 |     {
2398 |       "cell_type": "code",
2399 |       "metadata": {
2400 |         "colab": {
2401 |           "base_uri": "https://localhost:8080/"
2402 |         },
2403 |         "id": "zVoMyCBdnCMl",
2404 |         "outputId": "6e03c07f-a2b8-4ade-daa0-fab8d9511bad"
2405 |       },
2406 |       "source": [
2407 |         "# calculate AUC\n",
2408 |         "metrics.roc_auc_score(y_test, y_pred_prob)"
2409 |       ],
2410 |       "execution_count": 62,
2411 |       "outputs": [
2412 |         {
2413 |           "output_type": "execute_result",
2414 |           "data": {
2415 |             "text/plain": [
2416 |               "0.9835714940001832"
2417 |             ]
2418 |           },
2419 |           "metadata": {
2420 |             "tags": []
2421 |           },
2422 |           "execution_count": 62
2423 |         }
2424 |       ]
2425 |     },
2426 |     {
2427 |       "cell_type": "markdown",
2428 |       "metadata": {
2429 |         "id": "g8EoAyq6npFn"
2430 |       },
2431 |       "source": [
2432 |         "## Conclusion:\n",
2433 |         "\n",
2434 |         "### Thus the model is 98% accurate and most spam messeages have been categorized accordingly."
2435 |       ]
2436 |     },
2437 |     {
2438 |       "cell_type": "code",
2439 |       "metadata": {
2440 |         "id": "fyp0b_8fngFu"
2441 |       },
2442 |       "source": [
2443 |         ""
2444 |       ],
2445 |       "execution_count": null,
2446 |       "outputs": []
2447 |     }
2448 |   ]
2449 | }


--------------------------------------------------------------------------------