unit_8_py/lab-cross-validation/imb_solution.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "03392f2c-4553-42a3-b014-e5a8d95f7571",
   "metadata": {},
   "source": [
    "![logo_ironhack_blue 7](https://user-images.githubusercontent.com/23629340/40541063-a07a0a8a-601a-11e8-91b5-2f13e4e6b441.png)\n",
    "\n",
    "# Lab | Cross Validation\n",
    "\n",
    "For this lab, we will build a model on customer churn binary classification problem. You will be using `files_for_lab/Customer-Churn.csv` file."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95230d07-28fd-4bc2-92a9-96470f5674a7",
   "metadata": {},
   "source": [
    "<h1 style=\"color: #00BFFF;\">00 |</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2f44ccbc-7f45-4dbe-9342-3dcd05e9fe75",
   "metadata": {},
   "source": [
    "- Import the required libraries and modules that you would need."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5dfc6d25-c532-434f-a03f-cca19a85437e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 📚 Basic libraries\n",
    "import pandas as pd # data manipulation\n",
    "import numpy as np # numerical operations\n",
    "import matplotlib.pyplot as plt # 2D visualizations\n",
    "import warnings # warning messages management\n",
    "\n",
    "# 🤖 Machine Learning\n",
    "from sklearn.preprocessing import MinMaxScaler # data scaler\n",
    "from sklearn.tree import DecisionTreeClassifier # New ! Decision Tree\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split # train/test sets\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_curve, confusion_matrix, ConfusionMatrixDisplay # roc curve, confussion matrix & displayer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fe4fdb09-1050-4323-9397-9a46039717ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ⚙️ Settings\n",
    "pd.set_option('display.max_columns', None) # display all columns\n",
    "warnings.filterwarnings('ignore') # ignore warnings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "893d628a-e0ee-437b-9525-026341e087f0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 🔧 Basic functions\n",
    "def explore_data(data): # sum & returns duplicates, NaN & empty spaces\n",
    "    duplicate_rows = data.duplicated().sum()\n",
    "    nan_values = data.isna().sum()\n",
    "    empty_spaces = data.eq(' ').sum()\n",
    "    import pandas as pd\n",
    "    exploration = pd.DataFrame({\"NaN\": nan_values, \"EmptySpaces\": empty_spaces}) # New dataframe with the results\n",
    "    print(f\"There are {data.duplicated().sum()} duplicate rows. Also;\")\n",
    "    return exploration"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74f5a303-1b9d-4fda-affa-1ffe5bbd1393",
   "metadata": {},
   "source": [
    "<h1 style=\"color: #00BFFF;\">00 | Logistic Regression</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d2af6c11-6201-49b1-a891-52dc79aca698",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">01 | Data Extraction</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d8b3fa47-f7e7-4033-b428-ddf26df59ad1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gender</th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>Partner</th>\n",
       "      <th>Dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>PhoneService</th>\n",
       "      <th>OnlineSecurity</th>\n",
       "      <th>OnlineBackup</th>\n",
       "      <th>DeviceProtection</th>\n",
       "      <th>TechSupport</th>\n",
       "      <th>StreamingTV</th>\n",
       "      <th>StreamingMovies</th>\n",
       "      <th>Contract</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "      <th>Churn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>29.85</td>\n",
       "      <td>29.85</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>34</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>56.95</td>\n",
       "      <td>1889.5</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>2</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>53.85</td>\n",
       "      <td>108.15</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>45</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>42.30</td>\n",
       "      <td>1840.75</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>2</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>70.70</td>\n",
       "      <td>151.65</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender  SeniorCitizen Partner Dependents  tenure PhoneService  \\\n",
       "0  Female              0     Yes         No       1           No   \n",
       "1    Male              0      No         No      34          Yes   \n",
       "2    Male              0      No         No       2          Yes   \n",
       "3    Male              0      No         No      45           No   \n",
       "4  Female              0      No         No       2          Yes   \n",
       "\n",
       "  OnlineSecurity OnlineBackup DeviceProtection TechSupport StreamingTV  \\\n",
       "0             No          Yes               No          No          No   \n",
       "1            Yes           No              Yes          No          No   \n",
       "2            Yes          Yes               No          No          No   \n",
       "3            Yes           No              Yes         Yes          No   \n",
       "4             No           No               No          No          No   \n",
       "\n",
       "  StreamingMovies        Contract  MonthlyCharges TotalCharges Churn  \n",
       "0              No  Month-to-month           29.85        29.85    No  \n",
       "1              No        One year           56.95       1889.5    No  \n",
       "2              No  Month-to-month           53.85       108.15   Yes  \n",
       "3              No        One year           42.30      1840.75    No  \n",
       "4              No  Month-to-month           70.70       151.65   Yes  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "churnData = pd.read_csv('Customer-Churn.csv')\n",
    "churnData.head(5) # call of the churnData"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1ef5c41-52e9-4264-aad3-2364bdeb87d4",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">02 | EDA</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c43d2328-d052-40e7-aa99-b79645bb300e",
   "metadata": {},
   "outputs": [],
   "source": [
    "churnData['TotalCharges'] = pd.to_numeric(churnData['TotalCharges'], errors='coerce')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9e8b56e-8c0d-4323-b782-e6f5c31df992",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">03 | Data Cleaning</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6b54cc13-2468-49b7-aedf-ae4a2dbe8289",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "49"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "churnData.duplicated().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "26e65a7c-f696-4dc2-980b-23c96678fa09",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's drop the duplicates\n",
    "churnData = churnData.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "06c0901b-2ce3-4cca-8aa5-df04e4ca13c9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "churnData.duplicated().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2c0d41f3-5c32-4468-b5b3-9d4806ed3b9f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender              0\n",
       "SeniorCitizen       0\n",
       "Partner             0\n",
       "Dependents          0\n",
       "tenure              0\n",
       "PhoneService        0\n",
       "OnlineSecurity      0\n",
       "OnlineBackup        0\n",
       "DeviceProtection    0\n",
       "TechSupport         0\n",
       "StreamingTV         0\n",
       "StreamingMovies     0\n",
       "Contract            0\n",
       "MonthlyCharges      0\n",
       "TotalCharges        0\n",
       "Churn               0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# So... apart from nulls, duplicates, what else we can check?\n",
    "churnData.eq(' ').sum() # empty spaces ofc, eq = an empty space & sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b368dfe7-93da-45e2-b8f5-a439b7505ea7",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Muahahaha, there are ofc some empty spaces we are not gonna miss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "7bc94790-7fe8-4f01-bc7c-1e686cbc4aed",
   "metadata": {},
   "outputs": [],
   "source": [
    "churnData['TotalCharges'] = churnData['TotalCharges'].replace(' ', np.nan)  # replacing it with NaN \n",
    "churnData = churnData.dropna(subset=['TotalCharges']) # and dropping them like NaN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f28ea449-2709-4022-aea9-33da1d6cfc16",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gender              0\n",
       "SeniorCitizen       0\n",
       "Partner             0\n",
       "Dependents          0\n",
       "tenure              0\n",
       "PhoneService        0\n",
       "OnlineSecurity      0\n",
       "OnlineBackup        0\n",
       "DeviceProtection    0\n",
       "TechSupport         0\n",
       "StreamingTV         0\n",
       "StreamingMovies     0\n",
       "Contract            0\n",
       "MonthlyCharges      0\n",
       "TotalCharges        0\n",
       "Churn               0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "churnData.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fff6464-87bd-4d3b-99b3-81a3aab64030",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">04 | Data Pre-Processing</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "d02af493-41c3-4b9f-839a-36186a972ae9",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = churnData[['tenure', 'SeniorCitizen', 'MonthlyCharges', 'TotalCharges']] # Selecting the features from the lab\n",
    "y = churnData['Churn'] # Is our customer churn or not?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d4b1e850-7be1-4569-8bf8-a154c6c0396a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# StandardScaler dosen't make any sense, since it's not normal distributed.\n",
    "# We will use minmax, even to seniorcitizen (it will not affect the model)\n",
    "scaler = MinMaxScaler()\n",
    "X_scaled = scaler.fit_transform(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f165bf6f-5208-4cd0-944b-f7d4bdb86ae1",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">05 | Modeling</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6a3da3e3-ad9e-4d9e-8272-16ec7626c8d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "5539e5b7-8200-442a-986c-d914715aee37",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression()"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = LogisticRegression() # We pick our model\n",
    "model.fit(X_train,y_train) # We fit the training data into the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a8bd17d0-761f-4c5b-ac97-01c380dfd906",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "672dd1f3-1517-4e22-8b94-0b092ec5c06b",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">06 | Reporting</h1>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "ff5e99e4-0160-45d5-aba5-ea80844b4291",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.83      0.91      0.87      1559\n",
      "         Yes       0.62      0.44      0.52       536\n",
      "\n",
      "    accuracy                           0.79      2095\n",
      "   macro avg       0.73      0.68      0.69      2095\n",
      "weighted avg       0.77      0.79      0.78      2095\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(y_test, predictions)\n",
    "\n",
    "# Plot the confusion matrix\n",
    "disp = ConfusionMatrixDisplay(confusion_matrix=cm)\n",
    "disp.plot(cmap='Blues')\n",
    "plt.show()\n",
    "\n",
    "print(\"\")\n",
    "print(classification_report(y_test, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67d7970b-6d92-4c4f-9eaa-fbcfda382978",
   "metadata": {
    "tags": []
   },
   "source": [
    "<h1 style=\"color: #00BFFF;\">00 | SMOTE</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b3689b9-8c0d-4ba5-8a69-4f95da0078c0",
   "metadata": {},
   "source": [
    "1. Apply SMOTE for upsampling the data\n",
    "\n",
    "    - Use logistic regression to fit the model and compute the accuracy of the model.\n",
    "    - Use decision tree classifier to fit the model and compute the accuracy of the model.\n",
    "    - Compare the accuracies of the two models."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71dce860-2035-4c79-a89d-6bcf248fce2a",
   "metadata": {},
   "source": [
    "<h3 style=\"color: #008080;\">SMOTE</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4ed1e179-d11e-4a6e-96b7-64839fb57fab",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.over_sampling import SMOTE # Oversampling librarie technique"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "bb46f9dc-533e-47cc-bc0f-a8e4e7ca7be3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Selecting our library\n",
    "smote = SMOTE()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "42627351-a106-4b9c-9030-61a2543e9ba1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>tenure</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>29.85</td>\n",
       "      <td>29.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>34</td>\n",
       "      <td>56.95</td>\n",
       "      <td>1889.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>53.85</td>\n",
       "      <td>108.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>45</td>\n",
       "      <td>42.30</td>\n",
       "      <td>1840.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>70.70</td>\n",
       "      <td>151.65</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SeniorCitizen  tenure  MonthlyCharges  TotalCharges\n",
       "0              0       1           29.85         29.85\n",
       "1              0      34           56.95       1889.50\n",
       "2              0       2           53.85        108.15\n",
       "3              0      45           42.30       1840.75\n",
       "4              0       2           70.70        151.65"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = churnData['Churn']\n",
    "X = churnData.select_dtypes(include=[np.number]) # Selecting Numericals\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "fa26c29e-c1c1-48e7-a290-3e70360fdaea",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalazing the data\n",
    "transformer = StandardScaler().fit(X)\n",
    "\n",
    "# Transformation\n",
    "X = transformer.transform(X)\n",
    "\n",
    "# Using RandomUnderSampler\n",
    "X_sm, y_sm = smote.fit_resample(X, y)\n",
    "\n",
    "# We define train and test for X and Y\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_sm, y_sm, test_size=0.3, random_state=100)\n",
    "\n",
    "# test_size = We give 30% for testing and 70% for testing\n",
    "# random_state = it'll improve the model to divide always the model in the same way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "e9fca0ce-cbbc-45ed-84b4-f3df64f001ce",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No     5133\n",
       "Yes    1850\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "310c0e53-3e70-4a21-9138-4ca4395db860",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Yes    5133\n",
       "No     5133\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_sm.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a9879889-fbcd-466f-aaa4-253bba9fa981",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7301948051948052"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We traing the training dataset using Logistic Regression\n",
    "logistic = LogisticRegression(random_state=0).fit(X_train, y_train)\n",
    "\n",
    "# Predictions\n",
    "predictions = logistic.predict(X_test)\n",
    "\n",
    "logistic.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "d10df908-94fa-4015-b7fe-07663180e79c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.74      0.73      0.73      1564\n",
      "         Yes       0.72      0.73      0.73      1516\n",
      "\n",
      "    accuracy                           0.73      3080\n",
      "   macro avg       0.73      0.73      0.73      3080\n",
      "weighted avg       0.73      0.73      0.73      3080\n",
      "\n"
     ]
    }
   ],
   "source": [
    "logistic_5 = classification_report(y_test, predictions)\n",
    "print(logistic_5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "b4ad849e-ba3c-4549-ab63-5cf889affd31",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression Classification Report:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.74      0.73      0.73      1564\n",
      "         Yes       0.72      0.73      0.73      1516\n",
      "\n",
      "    accuracy                           0.73      3080\n",
      "   macro avg       0.73      0.73      0.73      3080\n",
      "weighted avg       0.73      0.73      0.73      3080\n",
      "\n",
      "Decision Tree Classification Report:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.75      0.74      0.75      1564\n",
      "         Yes       0.74      0.75      0.74      1516\n",
      "\n",
      "    accuracy                           0.75      3080\n",
      "   macro avg       0.75      0.75      0.75      3080\n",
      "weighted avg       0.75      0.75      0.75      3080\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Logistic Regression\n",
    "lr = LogisticRegression()\n",
    "lr.fit(X_train, y_train)\n",
    "lr_pred = lr.predict(X_test)\n",
    "lr_report = classification_report(y_test, lr_pred)\n",
    "print('Logistic Regression Classification Report:')\n",
    "print(lr_report)\n",
    "\n",
    "# Decision Tree\n",
    "dt = DecisionTreeClassifier()\n",
    "dt.fit(X_train, y_train)\n",
    "dt_pred = dt.predict(X_test)\n",
    "dt_report = classification_report(y_test, dt_pred)\n",
    "print('Decision Tree Classification Report:')\n",
    "print(dt_report)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c52664af-3f6a-4567-b40d-630eb085fedd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Decision Tree Classification is slightly better, 1% for each class, with SMOTE. let's see for TomeLinks"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4ee7edf-dd50-4c24-a67c-6f4d2c0f8ccf",
   "metadata": {},
   "source": [
    "<h1 style=\"color: #00BFFF;\">00 | TomeLinks</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d71de5b9-a77c-447e-9008-7c258ca66609",
   "metadata": {},
   "source": [
    "2. Apply TomekLinks for downsampling\n",
    "\n",
    "    - It is important to remember that it does not make the two classes equal but only removes the points from the majority class that are close to other points in minority class.\n",
    "    - Use logistic regression to fit the model and compute the accuracy of the model.\n",
    "    - Use decision tree classifier to fit the model and compute the accuracy of the model.\n",
    "    - Compare the accuracies of the two models.\n",
    "    - You can also apply this algorithm one more time and check the how the imbalance in the two classes changed from the last time.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e912d06-1714-495d-9a3a-04fdbf13d869",
   "metadata": {},
   "source": [
    "<h3 style=\"color: #008080;\">TomeLinks</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f3738116-cd88-44d5-974a-931f58c2057c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from imblearn.under_sampling import TomekLinks # Undersampling librarie technique"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "b6a460ad-9469-409b-8ad0-2902a2dec171",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Selecting our library\n",
    "tl = TomekLinks(sampling_strategy='majority')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ffb7b9fa-a502-4547-942c-a2d8763abc24",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SeniorCitizen</th>\n",
       "      <th>tenure</th>\n",
       "      <th>MonthlyCharges</th>\n",
       "      <th>TotalCharges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>29.85</td>\n",
       "      <td>29.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>34</td>\n",
       "      <td>56.95</td>\n",
       "      <td>1889.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>53.85</td>\n",
       "      <td>108.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>45</td>\n",
       "      <td>42.30</td>\n",
       "      <td>1840.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>70.70</td>\n",
       "      <td>151.65</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SeniorCitizen  tenure  MonthlyCharges  TotalCharges\n",
       "0              0       1           29.85         29.85\n",
       "1              0      34           56.95       1889.50\n",
       "2              0       2           53.85        108.15\n",
       "3              0      45           42.30       1840.75\n",
       "4              0       2           70.70        151.65"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = churnData['Churn']\n",
    "X = churnData.select_dtypes(include=[np.number]) # Selecting Numericals\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "3ca4d35c-abaf-47d7-83d0-c2d59a2cf004",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalazing the data\n",
    "transformer = StandardScaler().fit(X)\n",
    "\n",
    "# Transformation\n",
    "X = transformer.transform(X)\n",
    "\n",
    "# Using RandomUnderSampler\n",
    "X_tl, y_tl = tl.fit_resample(X, y)\n",
    "\n",
    "# We define train and test for X and Y\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_tl, y_tl, test_size=0.3, random_state=100)\n",
    "\n",
    "# test_size = We give 30% for testing and 70% for testing\n",
    "# random_state = it'll improve the model to divide always the model in the same way"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "6c3a1463-fe44-433b-9227-18d591da18fb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No     5133\n",
       "Yes    1850\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b06b9fcf-781d-4830-af0d-b67cb4648586",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No     4623\n",
       "Yes    1850\n",
       "Name: Churn, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_tl.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "e2d334cb-0ab6-4343-bb72-549b29c2332b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7914521112255407"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We traing the training dataset using Logistic Regression\n",
    "logistic = LogisticRegression(random_state=0).fit(X_train, y_train)\n",
    "\n",
    "# Predictions\n",
    "predictions = logistic.predict(X_test)\n",
    "\n",
    "logistic.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "3e52c75b-8ab6-4bfc-b6e5-06987e3b8d72",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.83      0.89      0.86      1397\n",
      "         Yes       0.66      0.53      0.59       545\n",
      "\n",
      "    accuracy                           0.79      1942\n",
      "   macro avg       0.75      0.71      0.72      1942\n",
      "weighted avg       0.78      0.79      0.78      1942\n",
      "\n"
     ]
    }
   ],
   "source": [
    "logistic_6 = classification_report(y_test, predictions)\n",
    "print(logistic_6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "9dc40f7f-a5cb-4eb6-bd2f-71ea951e0d0a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression Classification Report after TomekLinks:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.83      0.89      0.86      1397\n",
      "         Yes       0.66      0.53      0.59       545\n",
      "\n",
      "    accuracy                           0.79      1942\n",
      "   macro avg       0.75      0.71      0.72      1942\n",
      "weighted avg       0.78      0.79      0.78      1942\n",
      "\n",
      "Decision Tree Classification Report after TomekLinks:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          No       0.83      0.83      0.83      1397\n",
      "         Yes       0.56      0.56      0.56       545\n",
      "\n",
      "    accuracy                           0.75      1942\n",
      "   macro avg       0.70      0.70      0.70      1942\n",
      "weighted avg       0.75      0.75      0.75      1942\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "# Logistic Regression\n",
    "lr_tl = LogisticRegression()\n",
    "lr_tl.fit(X_train, y_train)\n",
    "lr_tl_pred = lr_tl.predict(X_test)\n",
    "lr_tl_report = classification_report(y_test, lr_tl_pred)\n",
    "print('Logistic Regression Classification Report after TomekLinks:')\n",
    "print(lr_tl_report)\n",
    "\n",
    "# Decision Tree\n",
    "dt_tl = DecisionTreeClassifier()\n",
    "dt_tl.fit(X_train, y_train)\n",
    "dt_tl_pred = dt_tl.predict(X_test)\n",
    "dt_tl_report = classification_report(y_test, dt_tl_pred)\n",
    "print('Decision Tree Classification Report after TomekLinks:')\n",
    "print(dt_tl_report)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "0af697cb-08b7-42ba-ab6d-3330f06d9433",
   "metadata": {},
   "outputs": [],
   "source": [
    "# In case of Tomelinks, downsampling, we got worst results with the DecisionTree predicting the Yes"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}