Project_Unsupervised_Learning_Learner's_Notebook(1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JxEtxNhXKpQv"
   },
   "source": [
    "----------------------------------------\n",
    "\n",
    "# Project: Unsupervised Learning\n",
    "- **Marks: 35**\n",
    "\n",
    "-----------------------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tIOusyeT4_RJ"
   },
   "source": [
    "Dear Learner, \n",
    "\n",
    "Welcome to the USL project.\n",
    "\n",
    "For this exercise, we will be using the Credit Card Customer Data which has been provided separately.\n",
    "\n",
    "---------------------------\n",
    "## Background: \n",
    "-----------------------------\n",
    "AllLife Bank wants to focus on its credit card customer base in the next\n",
    "financial year. They have been advised by their marketing research team, that the\n",
    "penetration in the market can be improved. Based on this input, the Marketing team\n",
    "proposes to run personalised campaigns to target new customers as well as upsell to\n",
    "existing customers. Another insight from the market research was that the customers\n",
    "perceive the support services of the back poorly. Based on this, the Operations team\n",
    "wants to upgrade the service delivery model, to ensure that customers queries are\n",
    "resolved faster. Head of Marketing and Head of Delivery both decide to reach out to the\n",
    "Data Science team for help.\n",
    "Objective: To identify different segments in the existing customer based on their\n",
    "spending patterns as well as past interaction with the bank.\n",
    "\n",
    "--------------------------\n",
    "## About the data:\n",
    "-----------------\n",
    "The data has customers as a data points with the following features-\n",
    "- Sl_no - Customer Serial Number\n",
    "- Customer Key - Customer identification\n",
    "- Avg_Credit_Limit\t- Average credit limit(currency less)\n",
    "- Total_Credit_Cards\t- Total number of credit cards \n",
    "- Total_visits_bank\t- Total bank visits\n",
    "- Total_visits_online -\t Total online visits\n",
    "- Total_calls_made - Total calls made\n",
    "\n",
    "\n",
    "---------------------------\n",
    "## Guidelines\n",
    "-----------------------------------------\n",
    "- The exercise consists of semi written code blocks. You need to fill the blocks as per the instructions to achieve the required results.\n",
    "- To be able to complete the assessment in the expected time, do not change the variable names. The codes might throw errors when the names are changed. \n",
    "- The marks of each requirement is mentioned in the question.\n",
    "- You can raise your issues on the discussion forum on the Olympus.\n",
    "- You can use Jupyter Notebook/ Google Colab(recommended) for this assessment.\n",
    "- Uncomment the code snippets and work on them\n",
    "\n",
    "--------------------------------------------\n",
    "## Resources\n",
    "-------------------------------------\n",
    "\n",
    "- Kmeans documentation - https://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html\n",
    "- GMM documentation - https://scikit-learn.org/stable/modules/generated/sklearn.mixture.GaussianMixture.html\n",
    "- DBSCAN - https://scikit-learn.org/stable/modules/generated/sklearn.cluster.DBSCAN.html\n",
    "\n",
    "\n",
    "Wishing you all the best!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "KJKn2dkPKnkU"
   },
   "outputs": [],
   "source": [
    "#Import all the necessary packages\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pylab as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import seaborn as sns\n",
    "from sklearn.cluster import KMeans\n",
    "\n",
    "import sklearn\n",
    "import sklearn.metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "xUObRKKPYo2b"
   },
   "outputs": [],
   "source": [
    "# Just some fancy stuff to make make notebooks more presentable\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore') # to get rid of warning messages\n",
    "\n",
    "# Remove scientific notations and display numbers with 2 decimal points instead\n",
    "pd.options.display.float_format = '{:,.2f}'.format        \n",
    "\n",
    "# Update default background style of plots\n",
    "sns.set_style(style='darkgrid')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MeIShU1-hh0U"
   },
   "source": [
    "### 1. Read the data file (1 mark)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "u6UKbCjXhfDJ"
   },
   "outputs": [],
   "source": [
    "# uncomment below code\n",
    "df = pd.read_excel('Credit Card Customer Data.xlsx')\n",
    "\n",
    "# Note -- if you are using google colab you can use the mount drive feature(or any other alternative) but make sure that the dataframe in which the data file is read is named 'df'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "P1kDTEVUK7WI",
    "outputId": "7604e29c-b4a2-45bf-a87b-8bf348d6a83d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(660, 7)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "lMcUBOrbLCZM",
    "outputId": "75d45c7f-5e49-4636-e86c-3bc6f8b4c86d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 660 entries, 0 to 659\n",
      "Data columns (total 7 columns):\n",
      " #   Column               Non-Null Count  Dtype\n",
      "---  ------               --------------  -----\n",
      " 0   Sl_No                660 non-null    int64\n",
      " 1   Customer Key         660 non-null    int64\n",
      " 2   Avg_Credit_Limit     660 non-null    int64\n",
      " 3   Total_Credit_Cards   660 non-null    int64\n",
      " 4   Total_visits_bank    660 non-null    int64\n",
      " 5   Total_visits_online  660 non-null    int64\n",
      " 6   Total_calls_made     660 non-null    int64\n",
      "dtypes: int64(7)\n",
      "memory usage: 36.2 KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "PiOr4bspLM_v",
    "outputId": "60886f6f-4629-4679-c0a3-94c6c9b8bea1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check for the duplicate \n",
    "df.duplicated().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rjw576kPNzlS",
    "outputId": "1466e4ec-143b-4279-feb4-94a4f43bc25c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sl_No                  0\n",
       "Customer Key           0\n",
       "Avg_Credit_Limit       0\n",
       "Total_Credit_Cards     0\n",
       "Total_visits_bank      0\n",
       "Total_visits_online    0\n",
       "Total_calls_made       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# no duplicates, lets check for missing values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KWyGo_wWjkvX"
   },
   "source": [
    "### 2. Find the number of unique entries in each of the columns. What can be inferred from the observation about the the Customer key feature? (2 marks)\n",
    "\n",
    "- Unique entries (1 mark)\n",
    "- Inferences (1 mark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "Gn7zX2ZIN0iB"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sl_No                  660\n",
       "Customer Key           655\n",
       "Avg_Credit_Limit       110\n",
       "Total_Credit_Cards      10\n",
       "Total_visits_bank        6\n",
       "Total_visits_online     16\n",
       "Total_calls_made        11\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# No missing values. Let us now figure out the uniques in each column. \n",
    "\n",
    "\n",
    "# uncomment below code\n",
    "df.nunique()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ogwZ1VtJkciz"
   },
   "source": [
    "**Add inference from the above observation about the Customer Key feature here**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is 655 unique Customers , total numbers are 660 so it means 5 of them are duplicate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 197
    },
    "id": "qLuEIKrxOr3e",
    "outputId": "bd90c2f4-1b25-4deb-8406-47a5ec9b0edc"
   },
   "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>Sl_No</th>\n",
       "      <th>Customer Key</th>\n",
       "      <th>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>332</th>\n",
       "      <td>333</td>\n",
       "      <td>47437</td>\n",
       "      <td>17000</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>398</th>\n",
       "      <td>399</td>\n",
       "      <td>96929</td>\n",
       "      <td>67000</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>433</td>\n",
       "      <td>37252</td>\n",
       "      <td>59000</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>541</th>\n",
       "      <td>542</td>\n",
       "      <td>50706</td>\n",
       "      <td>60000</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>632</th>\n",
       "      <td>633</td>\n",
       "      <td>97935</td>\n",
       "      <td>187000</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Sl_No  Customer Key  Avg_Credit_Limit  Total_Credit_Cards  \\\n",
       "332    333         47437             17000                   7   \n",
       "398    399         96929             67000                   6   \n",
       "432    433         37252             59000                   6   \n",
       "541    542         50706             60000                   7   \n",
       "632    633         97935            187000                   7   \n",
       "\n",
       "     Total_visits_bank  Total_visits_online  Total_calls_made  \n",
       "332                  3                    1                 0  \n",
       "398                  2                    2                 2  \n",
       "432                  2                    1                 2  \n",
       "541                  5                    2                 2  \n",
       "632                  1                    7                 0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# There are some duplicates in customer key. Let us explore\n",
    "\n",
    "duplicate_keys = df['Customer Key'].duplicated()\n",
    "df[duplicate_keys]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vEjJ7d_FPiIr",
    "outputId": "2fa8cbc6-8adf-4c60-cb1c-aba4a7d33209"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 655 entries, 0 to 659\n",
      "Data columns (total 7 columns):\n",
      " #   Column               Non-Null Count  Dtype\n",
      "---  ------               --------------  -----\n",
      " 0   Sl_No                655 non-null    int64\n",
      " 1   Customer Key         655 non-null    int64\n",
      " 2   Avg_Credit_Limit     655 non-null    int64\n",
      " 3   Total_Credit_Cards   655 non-null    int64\n",
      " 4   Total_visits_bank    655 non-null    int64\n",
      " 5   Total_visits_online  655 non-null    int64\n",
      " 6   Total_calls_made     655 non-null    int64\n",
      "dtypes: int64(7)\n",
      "memory usage: 40.9 KB\n"
     ]
    }
   ],
   "source": [
    "# let us just drop them for simplicity.\n",
    "\n",
    "df = df[~duplicate_keys]\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jj7U8RQTlw49"
   },
   "source": [
    "### 3. Drop the columns Sl_No and Customer Key(1 mark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "EOoF3Jv1Pjkq"
   },
   "outputs": [],
   "source": [
    "# Great, we have done some basic checks. Now let us drop the variables that are not required\n",
    "\n",
    "# Uncomment below\n",
    "df.drop(columns = ['Sl_No', 'Customer Key'], inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 137
    },
    "id": "4cL9VEPCSVqX",
    "outputId": "e1312b2a-ae29-4616-f873-8924b7d133ec"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50000</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50000</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "0            100000                   2                  1   \n",
       "1             50000                   3                  0   \n",
       "2             50000                   7                  1   \n",
       "\n",
       "   Total_visits_online  Total_calls_made  \n",
       "0                    1                 0  \n",
       "1                   10                 9  \n",
       "2                    3                 4  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 377
    },
    "id": "Xk0Ce0ZmSWc9",
    "outputId": "d75502d8-c22a-4527-ba41-153e0ca87621"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>8000</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>6000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>215</th>\n",
       "      <td>8000</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>10000</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>324</th>\n",
       "      <td>9000</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>361</th>\n",
       "      <td>18000</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>378</th>\n",
       "      <td>12000</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>385</th>\n",
       "      <td>8000</td>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>5000</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>455</th>\n",
       "      <td>47000</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>497</th>\n",
       "      <td>52000</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "162              8000                   2                  0   \n",
       "175              6000                   1                  0   \n",
       "215              8000                   4                  0   \n",
       "295             10000                   6                  4   \n",
       "324              9000                   4                  5   \n",
       "361             18000                   6                  3   \n",
       "378             12000                   6                  5   \n",
       "385              8000                   7                  4   \n",
       "395              5000                   4                  5   \n",
       "455             47000                   6                  2   \n",
       "497             52000                   4                  2   \n",
       "\n",
       "     Total_visits_online  Total_calls_made  \n",
       "162                    3                 4  \n",
       "175                    2                 5  \n",
       "215                    4                 7  \n",
       "295                    2                 3  \n",
       "324                    0                 4  \n",
       "361                    1                 4  \n",
       "378                    2                 1  \n",
       "385                    2                 0  \n",
       "395                    0                 1  \n",
       "455                    0                 4  \n",
       "497                    1                 2  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#let us check for duplicates now. Duplicates would mean customers with identical features\n",
    "\n",
    "df[df.duplicated()]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ogu8TnJjUrgL",
    "outputId": "1c46d588-00bb-4fc5-a307-4d6703084d22"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 644 entries, 0 to 659\n",
      "Data columns (total 5 columns):\n",
      " #   Column               Non-Null Count  Dtype\n",
      "---  ------               --------------  -----\n",
      " 0   Avg_Credit_Limit     644 non-null    int64\n",
      " 1   Total_Credit_Cards   644 non-null    int64\n",
      " 2   Total_visits_bank    644 non-null    int64\n",
      " 3   Total_visits_online  644 non-null    int64\n",
      " 4   Total_calls_made     644 non-null    int64\n",
      "dtypes: int64(5)\n",
      "memory usage: 30.2 KB\n"
     ]
    }
   ],
   "source": [
    "#let us drop these too\n",
    "df=df[~df.duplicated()]\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GWbxYcHCmjh6"
   },
   "source": [
    "### 4. Visualise through box plot  (5 marks)\n",
    "\n",
    "- Extract the features(column names) of the df in a list(1 mark)\n",
    "- Complete the function below that takes the above list of column names and the df and plots the boxplot for all the features(3 marks)\n",
    "- Call the function with the features and the df(1 mark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "SN02moxuWGf2",
    "outputId": "096f7b9b-1aac-42ed-9070-e455a0a49306"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Now lets go ahead with the exploring each variable at hand. We will use the boxplot for these\n",
    "\n",
    "\n",
    "features = list(df.columns)\n",
    "#features\n",
    "\n",
    "# write the function below. Iterate through the features plot the boxplot for all the featrues one by one.\n",
    "\n",
    "def boxplots_plot(features, df):\n",
    "   for i in features:\n",
    "     sns.boxplot(df[i])\n",
    "     plt.show()\n",
    "  \n",
    "\n",
    "boxplots_plot(features, df)\n",
    "\n",
    "# The plots are drawn for your reference."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "IkLtnnkcp0cl"
   },
   "source": [
    "### 5. Draw the correlation heatmap with annotation(4 marks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 368
    },
    "id": "ENObY0wEWhAD",
    "outputId": "19875a20-2190-401d-eb74-58ad35abb75a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fcbe3c34a90>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# We see that there are some outliers in the Average credit limit and total visits online features\n",
    "# Now let us generate a correlation heatmap\n",
    "sns.heatmap( df.corr(), annot = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "skjNQIiMq2Ar"
   },
   "source": [
    "### 6. Scale the dataset, create a dataframe for the scaled value and display the top 3 entries (5 marks)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 137
    },
    "id": "gtAGq41QYFvO",
    "outputId": "0b0c1602-8b73-4ac0-dd3c-ba97730f0308"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "0              0.49                0.11               0.20   \n",
       "1              0.24                0.22               0.00   \n",
       "2              0.24                0.67               0.20   \n",
       "\n",
       "   Total_visits_online  Total_calls_made  \n",
       "0                 0.07              0.00  \n",
       "1                 0.67              0.90  \n",
       "2                 0.20              0.40  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let us now scale the dataset and present the scaled values in a dataframe and display the top 3\n",
    "\n",
    "# use the fit tranform function on the original dataframe\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "mms = MinMaxScaler()\n",
    "df_scaled = mms.fit_transform(df)\n",
    "\n",
    "# creating the dataframe of scaled values\n",
    "df_scaled = pd.DataFrame(df_scaled)\n",
    "\n",
    "# use the stored feature names to define the column names for the new dataframe\n",
    "df_scaled.columns = features\n",
    "\n",
    "# print the top3 entries in the scaled data frame\n",
    "df_scaled.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 197
    },
    "id": "3vEUSvZLW9po",
    "outputId": "6f5b163a-2415-4c0e-d694-4eedc3c3f9b4"
   },
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Avg_Credit_Limit</th>\n",
       "      <td>644.00</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.23</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <td>644.00</td>\n",
       "      <td>0.41</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.56</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <td>644.00</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.80</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total_visits_online</th>\n",
       "      <td>644.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.27</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total_calls_made</th>\n",
       "      <td>644.00</td>\n",
       "      <td>0.36</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.53</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     count  mean  std  min  25%  50%  75%  max\n",
       "Avg_Credit_Limit    644.00  0.16 0.19 0.00 0.04 0.08 0.23 1.00\n",
       "Total_Credit_Cards  644.00  0.41 0.24 0.00 0.22 0.44 0.56 1.00\n",
       "Total_visits_bank   644.00  0.48 0.33 0.00 0.20 0.40 0.80 1.00\n",
       "Total_visits_online 644.00  0.17 0.20 0.00 0.07 0.13 0.27 1.00\n",
       "Total_calls_made    644.00  0.36 0.29 0.00 0.10 0.30 0.53 1.00"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_scaled.describe().T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "oPSv-4Hat5Wg"
   },
   "source": [
    "### 7. Elbow Plot(3 marks)\n",
    "\n",
    "#### K-Means clustering\n",
    "\n",
    "Now that we have scaled, let us move to KMeans clustering."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "id": "5_IDWbX6amaq"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAEOCAYAAACdLzzJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3de1gU9f4H8PfsLLuwoawmaaZ4K4+WdlTSNPG6WJpmZeb1YGYHjz5mZWYqJaKmZpla+DMU86hkmuXlWD3HLmjmeiEveK1jheYVUQEzgYXdmfn9MQGiiAK7zO7s+/U8PrsMuzMfwOc9s5/vd2YERVEUEBGRbhi0LoCIiNyLwU5EpDMMdiIinWGwExHpDIOdiEhnGOxERDrjVcF+8OBBREVF3fJ1J0+eRJ8+fW5Yvnz5csydO9cTpRER+Qyj1gUUSkxMxKZNmxAUFFTm6zZu3IiVK1ciOzu7aJnD4cCbb76JQ4cO4dFHH/V0qUREXs1rjtjDwsIQHx9f9PWxY8cQFRWFqKgojB07Fn/++ScAICQkBB9//HGJ9+bn5+Opp57CqFGjqrRmIiJv5DXB/thjj8FoLP4AMWXKFEydOhVJSUno3Lkzli5dCgDo1q0bLBZLifeGhIQgIiKiSuslIvJWXtOKuV5aWhqmTZsGAHA6nWjUqJHGFRER+QavDfZGjRphzpw5qFu3Lvbt24eLFy9qXRIRkU/w2mCPi4vDxIkTIUkSAGDmzJkaV0RE5BsEXt2RiEhfvGbwlIiI3MMrWjGyLEOSKvbBQRSFCr/Xk1hX+bCu8mFd5aPXugICxFKXe0WwS5KCy5dzK/Req9VS4fd6EusqH9ZVPqyrfPRaV2hotVKXsxVDRKQzDHYiIp1hsBMR6QyDnYhIZxjsREQ645PBHh9vgt1ecpqP3S4iPt6kUUVERN7DJ4O9dWsJ0dGBReFut4uIjg5E69aSxpUREWnPK+axl1dEhITERAeefz4ITz6p4KuvApGY6EBEBIOdiMgnj9gBNdzr15excqUBzz3nZKgTEf3FZ4Pdbhdx4oRa/kcf3dhzJyLyVz4Z7IU99Q8+cAAAevd2lui5ExH5M58M9tRUEYmJDjzxhAutWytISzMgMdGB1FQGOxGRTw6ejh1bUPS8Z08Fc+aIaNFCYp+diAg+esR+rV69FMiygO+/98l9FBGR2/l8sLdtC9SooeC77xjsRESADoJdFIFu3VzYskWELGtdDRGR9nw+2AHAZnPh0iUDDh7UxY9DRFQpukjCbt0kCIKC5GS2Y4iIdBHstWopaNNGZrATEUEnwQ6o7Zj9+w24dEnQuhQiIk3pJtgjI11QFAFbt/IkJSLyb7oJ9gcflFGrFtsxRES6CXaDAejeXcLWrUZIPAGViPyYboIdUNsx2dkC9u/X1Y9FRFQuukrArl1dMBg47ZGI/Juugt1qBdq2lXh5ASLya7oKdgCIjJRw6JCIjAxOeyQi/6S7YLfZXACALVs47ZGI/JNHgt3pdGL8+PEYNGgQhgwZgrS0NE9splQPPCCjTh2Z7Rgi8lseCfZt27bB5XJhzZo1GDNmDBYsWOCJzZRKENSj9m3bjHA6q2yzRERewyPB3qhRI0iSBFmWcfXqVRiNVXv0bLNJuHJFwN69bMcQkf/xSOJaLBacPXsWvXr1QnZ2NhISEsp8vSgKsFotFdqWKBpueG/fvoDRqGD79kD06qVUaL2VVVpd3oB1lQ/rKh/WVT6eqktQFMXtyTd79myYTCaMHz8e6enpeO655/DFF1/AbDaX+nqnU8Lly7kV2pbVain1vf36BSEzU8C2bRVbb2XdrC6tsa7yYV3lw7rKp7J1hYZWK3W5R1ox1atXR7Vq6gZDQkLgcrkgVfF5/jabCz//LOLsWU57JCL/4pFgHz58OI4ePYohQ4bgueeew7hx42CxVO3HIJtN3ZHwLFQi8jceSb077rgD77//vidWfdv+9jcZ9erJSE4WMWwYp8cQkf/Q3QlKhQqnPf7wgxH5+VpXQ0RUdXQb7IB6tcecHAEpKZz2SET+Q9fBHhEhwWRSeBYqEfkVXQf7HXcAHTpISE7mETsR+Q9dBzugtmN+/VXEyZOc9khE/sEvgh3gtEci8h+6D/bGjRU0bMibXBOR/9B9sAuCetRut4vIy9O6GiIiz9N9sANqsOflCdi1i4OoRKR/fhHsHTpICAzktEci8g9+EexBQeqcdvbZicgf+EWwA+rlBU6cMOD4cU57JCJ986tgB8B2DBHpnt8Ee8OGCu67T2KwE5Hu+U2wA+o12nfuFJGTo3UlRESe42fB7kJBgQC7ndMeiUi//CrY27eXYLEonB1DRLrmV8FuNgOdO7uQnGyE+2/hTUTkHfwq2AEgMlLC6dMG/PKL3/3oROQn/C7diqc9ss9ORPrkd8F+zz0KmjfnWahEpF9+F+yAetS+e7eIP//UuhIiIvfzy2CPjJTgcgn44QcetROR/vhlsLdtK6FaNYX3QiUiXfLLYA8IALp2deG77zjtkYj0xy+DHVBvvnH+vAFHj/rtr4CIdMpvU617dwkAb3JNRPrjt8Feu7aCli0lzmcnIt3x22AH1HbM3r0iLl/WuhIiIvfx62C32VyQJAHbtrEdQ0T64dfBHh4uo0YN3uSaiPTFr4NdFIFu3VxIThYhy1pXQ0TkHn4d7ADQvbsLly4ZcOiQ3/8qiEgn/D7NunWTIAhsxxCRfvh9sIeGKmjdWuZ8diLSDb8PdkCdHbN/vwGZmYLWpRARVZrHgn3x4sUYOHAg+vXrh88++8xTm3GLyEgXFEXA1q08WYmIfJ9Hgj0lJQWpqalYvXo1kpKScP78eU9sxm3+/ncZtWrJ7LMTkS54JMnsdjuaNm2KMWPG4OrVq3j99dc9sRm3MRjUQdTvvjNCktRpkEREvsojwZ6dnY1z584hISEBZ86cwejRo7F582YIQuk9bFEUYLVaKrQtUTRU+L3X6ttXwGefCfj1Vwvat6/06txWl7uxrvJhXeXDusrHU3V5JNitVisaN24Mk8mExo0bw2w2IysrC3feeWepr5ckBZcv51ZwW5YKv/da7doBBkMwNm50oVmzgkqvz111uRvrKh/WVT6sq3wqW1doaLVSl3ukxx4eHo7t27dDURRkZGQgLy8PVqvVE5tymxo1gIce4k2uicj3eSTFunXrhj179qB///5QFAWxsbEQfaBxHRkpYdYsMzIyBNSuzVsrEZFv8tjhqbcPmJbGZnNh1iwztm4VMWiQS+tyiIgqhCcoXaNFCxm1a3PaIxH5Ngb7NQRBPWr//nsjXDxgJyIfxWC/js0m4coVAXv3ev+YABFRaRjs1+nSxQWjUeG9UInIZzHYr1O9OvDwwxL77ETksxjspejeXcJPP4k4d45XeyQi38NgL0VkpDpyypOViMgXMdhL0ayZjHvukdlnJyKfxGAvReG0xx9+MKKg8peNISKqUgz2m4iMdCEnR0BKCo/aici3MNhvIiJCgsnEm1wTke9hsN9EcDDQvr2E5GQesRORb2GwlyEy0oVffhFx8iSnPRKR72Cwl4HTHonIFzHYy9CkiYIGDWRs2cJgJyLfwWAvgyCoR+3bt4twOLSuhojo9jDYbyEy0oW8PAE7d3IQlYh8A4P9Fh55REJgoMI+OxH5DAb7LQQFAR078mqPROQ7GOy3ITLShRMnDDh+nNMeicj7lRnsFy5cKHX5wYMHPVKMt7LZOO2RiHxHmcH+2muvFT2fMGFC0fP33nvPcxV5oYYNFdx7L9sxROQbygx2RVGKnp8/f77U5f7CZpOwc6eInBytKyEiKluFeuyC4H+9ZpvNhfx8ATt2cNojEXm3MoP92gD3xzC/VocOEiwWXu2RiLxfmSm1f/9+REREAAAuX75c9PyPP/7wfGVexmwGOnd2ITnZCEXJh5/v54jIi5UZ7EeOHKmqOnyCzSZh8+YA/PqrAU2bylqXQ0RUqjJbMQUFBVixYgUURUFGRgZeeuklvPbaa7h48WJV1edVCqc98l6oROTNygz2GTNm4Ny5c5BlGXFxcWjWrBkee+wxxMXFVVF53qVePQXNmkmcz05EXq3MYD937hwmT54Ml8uFffv2ITo6Gj169EBWVlZV1ed1bDYJu3eL+PNPrSshIirdbc2K2b9/P1q2bImAgAAAQH5+vucr81KRkS44nQJ++IFH7UTkncoMdovFgk8//RSJiYno3bs3ZFnGZ599hrvvvruq6vM67dpJCA5WeC9UIvJaZQZ7XFwcTp06BZvNhn79+iE5ORlffvml3/bYASAgAOjatXDao9bVEBHdqMxgT09Ph91ux4ABA/DNN99g6tSpOH/+PA4fPlxV9XmlyEgX0tMN+OknXhyTiLxPmck0f/58zJkzBwEBAViwYAESExOxbt06LFmypKrq80rdu0sAeLVHIvJOt7wIWLNmzZCRkYG8vDw88MADCA4OhsHg30eqdeooaNFC4nx2IvJKZSa0LKtnV27fvh0dOnQAoJ60lMNLHCIy0oU9e0Rcvqx1JUREJZUZ7B06dMCgQYOwcOFCREVF4dSpUxg1ahQef/zxW644MzMTXbp0QVpamtuK9SY2mwRJErBtG9sxRORdykylkSNHwmazoWbNmqhRowZOnTqFwYMHo0ePHmWu1Ol0IjY2FoGBgW4t1puEh0uwWtWbXD/5pEvrcoiIigiKB+6a8dZbb6FLly5YsmQJ4uLi0KRJkzJfL8syJKliZYiiAZKkzQW5/vEPAdu2CTh5Usb1ww5a1lUW1lU+rKt8WFf5VLaugIDSx/nc3kdYv349atasiU6dOt327BlJUnD5cm6Ftme1Wir83srq1MmItWuDsH27A3//e8k/jpZ1lYV1lQ/rKh/WVT6VrSs0tFqpy90+vWXdunXYuXMnoqKi8PPPP2PixIm6vRpk4bRH3nyDiLyJ2xNp1apVRc+joqIQFxeH0NBQd2/GK4SGKmjdWr3J9fjxBVqXQ0QEwANH7P7GZnNh/34DMjN5SyUi8g4eDfakpKRbDpz6ushIFxRFwPff82QlIvIOPGKvpFatZNSqJbPPTkReg8FeSQYD0LWrhK1bRUiS1tUQETHY3SIy0oWsLANSU/nrJCLtMYncoGtXFwwGhe0YIvIKDHY3qFkTCA+XsWULg52ItMdgd5PISBcOHBBx4QKnPRKRthjsbmKzqRcC27KF0x6JSFsMdjdp0ULGXXfJvKsSEWmOwe4mBoN6jfatW41w8Sq+RKQhBrsbRUa6cOWKgL172Y4hIu0w2N2oSxcXRFFBcjKDnYi0w2B3o+rVgYcfljifnYg0xWB3o/h4Exo3lnH0qIizZ9VldruI+HiTtoURkV9hsLtR69YSvvgiAADw9dcC7HYR0dGBaN2aF5EhoqrDYHejiAgJy5blwWBQ8N57AqKjA5GY6EBEBIOdiKoOg93NOnWSEB4u4ddfBTzyiMRQJ6Iqx2B3M7tdRFqaAffco+CLL4xYtYoDqURUtRjsblTYU1+61IEdO2RYrQrGjw/Ef//L6Y9EVHUY7G6UmioW9dTr1gU+/jgPggBMm2bmTTiIqMow2N1o7NiCEj31du1kvPNOPo4fFzFzJqc8ElHVYAPYw6KinDh0yICFC8148EEZTz3FC8kQkWfxiL0KzJyZj3btXHj55UAcOcJfORF5FlOmCphMwEcfORASomD48CBkZWldERHpGYO9itSureDf/87D+fMCoqODeGlfIvIYBnsVCg+X8e67DmzfbsSMGWatyyEineLgaRUbPNiFgwcL8OGHJjz4oIRnnuGhOxG5F4/YNTBjRj46dHBh3LhAHDrEPwERuRdTRQMBAcDSpQ7UrKkOpl66JGhdEhHpCINdI6GhCpYvz8PFiwJGjgyE06l1RUSkFwx2DbVqJWPuXAfsdiOmTeNgKhG5BwdPNTZwoAuHDxdgyRITWraUMHAgB1OJqHJ4xO4Fpk7NR8eOLrz2WiAOHOCfhIgqhyniBQICgMREB0JD1cHUCxc4mEpEFcdg9xK1aqmDqVlZ6i31OJhKRBXFYPciDz4oY948B3btMiI2loOpRFQxHDz1Mv37u3DoUAESEtQzUwcP5mAqEZWP24Pd6XQiJiYGZ8+eRUFBAUaPHg2bzebuzehabGw+jh41YMKEQDRtmovwcFnrkojIh7i9FbNp0yZYrVZ88sknSExMxIwZM9y9Cd0zGoHExDzUqaPg+eeDkJHBwVQiun1uD/aePXvi5ZdfLvpaFHkj54qoWRNYvjwPf/wh4IUXAlFQoHVFROQrBEVRFE+s+OrVqxg9ejQGDBiAJ554oszXyrIMSapYGaJogCR5X6vCXXV9+qmAqCgDRo6UsXBh5f9Uev99uRvrKh/WVT6VrSsgoPQDZ48Mnqanp2PMmDEYMmTILUMdACRJweXLuRXaltVqqfB7PclddT32GPDiiyYsXGjG3/7mQFRU5eZB6v335W6sq3xYV/lUtq7Q0GqlLnd7sF+6dAkjRoxAbGwsOnTo4O7V+6U33ijAkSMiJk0yo1kzCW3bet+RBxF5D7f32BMSEnDlyhUsWrQIUVFRiIqKgsPhcPdm/IooAosX56FuXQUjRgTh/HkOphLRzXmsx14eTqfEVsxt+OknAx5/3ILmzWVs3JgLcwXOYfKn35c7sK7yYV3l46lWDM889SH33y8jPt6BfftExMTwzFQiKh2D3cc88YQLL7+cj6QkE1asCNC6HCLyQgx2HzRpUgFsNhdiYsxISeF5AkRUEoPdB4kikJCQh3r1FIwYEYj0dA6mElExBruPCgkBVqzIQ26ugOefDwInHhFRIQa7D2vWTMbChQ7s36/Ocdd+fhMReQMGu4/r3duFV1/NxyefmLBsGQdTiYjBrguvv16ARx91YcoUM3bt4mAqkb9jsOuAwQAsWpSHhg1lvPBCIM6e5WAqkT9jsOtE9erAihUOOBwChg8PQl6e1hURkVYY7Dpy330yFi3Kw8GDIiZMCORgKpGfYrDrTM+eEiZMyMfatQFYupSDqUT+iMGuQ+PHF6BnTydiY82w2zmYSuRvGOw6ZDAA//d/DjRuLCM6OhCnT3MwlcifMNh1qlo1YOXKPBQUqIOpud53xVIi8hAGu441aaIgISEPR44YMH48B1OJ/AWDXed69JDQtauEdesCkJBQPJhqt4uIjzdpWBkReQqD3Q+8+GIBTCYFcXFmbN4MbN8uIjo6EK1bS1qXRkQewGD3A506Sfj3v/NgMAB9+4ro3z8IzZrJOHbMgCNHDJCY70S6YtS6AKoaPXpIGDnSiQ8/NKFpUxnHjxsweXIgAKB6dQVt20po317Cww9LaNVKQmCgxgUTUYUx2P2E3S5i7VojYmJkLF4sYMkSBxo0kLF7t4iUFPVfcrJ6H1WTSUHr1mrIt28voW1bCSEhGv8ARHTbGOx+wG5Xe+qJiQ706WPGQw/lF309YIALAwa4AACZmQJ+/FHE7t0ifvxRxKJFJnzwgQBBUNC8uVx0RN++vYS77+YUGyJvxWD3A6mpIhITHYiIUJvpERESEhMdSE0Vi5YBwJ13KujVy4VevdSgz8kB9u9Xj+Z37xaxZk0Ali1TZ9KEhclFIf/wwxLuu0+GwPOgiLyCoCjaz252OiVcvlyxM2isVkuF3+tJeqzL5QKOHDEUBX1KiohLl9Tx9zvvlNGuXfERfcuWMgLKcakaPf6+PIl1lY9e6woNrVbqch6x020zGoFWrWS0aiXjX/9yQlGA48cF7N5tLAr7//5XTXOLRUF4uBr0Dz8sITxcQnCwxj8AkZ9gsFOFCYJ6dmuTJk4MHeoEAJw/X9ynT0kRMW+eCbIsQBQVtGxZ3L5p107CmjUBaN1aKtEOsttFpKaKGDu2QKsfi8jnMdjJrerUUdC3rwt9+6p9+itXgL17i/v0y5cHYPFitU9ft66EzEwToqMLEB0N/PSTiBdfVAd1iaji2GP3ENZVuvx84OBBA1JS1PbNjh0icnKKR12tVhlNmigIC5NRv76M+vUV1K8vIyxMRr16SpXPr9f693UzrKt89FoXe+zkFcxmoF07Ge3aFWDsWECWgddfN2PlShM6dnShUSMZp04ZcOCAiC+/NMLpLDnV5q671LAPC5P/Cv/i4L/nnqoPfiJvxGAnTe3cKeKrrwpPnDJg/PiCop67JAEXLgg4edKA06cFnD6tPp46ZUBqaunBX7t2cfBff8R/u8EfH29i7598GoOdNFPWiVMRERJEEbj7bgV33y2hffsb3y9JQEaGGvSFwX/qlPq4b5+ITZuMcLlKBn+dOiXD/vrgN5uB1q2la+oqWSeRL2Cwk2Zu98SpmxFFoG5dBXXr3jz4z58vGfiFO4BbBf/998sYOjQInTsDdnsQRo8ugMMB7NghIihIgcWCEo9BQWo9nsRPEnS7OHjqIayrfLSoy+UqLfiL2z2nTwtQlNs/nTYwUA14i0UpCnv1ecnH0nYKhd8vXK6+ruSyPXtEjBxZ/Annyy9LfsLRyrU7nMK/o7ftcPT6/56Dp0TXMRqBevUU1KsnoUOHkt+z20X885+BiIpSsGKFgKlTHWjeXEZuroC8PJR4zM0F8vKEomXq8+LHP/8UkJEh3PAeWS7fNRgMBgUmE/DMM0GoXh24ejUIDRrIePttEwIDUbQTCAxUdzLqshu/Dgoq/vr611y7DuNtpoO3tq78+RMOg53oOoXBtHSpemTcpYvD7UfGigI4nSixAyi50yjeOVy7I8nJEbBrl4hDh0Tcd5+EBg0UOBzq8sxMwOEwwOEAHI7i95bnU8e1jMaSgX/tY+GOpPB5eLiEIUOC0KYNkJoahP79nfjpJwN++80As1ndIZnNuMnz0r9f2daWP+9wGOxE16ls7/92CAJgMqn/rNbCbuitu6J2u4jPPy++/PKcOfll1qQoQEEB/gp7NegdDqHE13l5QomdQeH3Ch9zc0t+7XCoO5qsrMKv1fW4XMDOnepOJCmp8rddFEXlr7BXLyVd+GgyqZ80Cp9f+/3rX9u9uwtDhwbhkUeAXbuCMGSIE2lpautNFNVPJUYj/nquFH1dvLzsZYXrCAhQSqzPYMBNL4pXFTscj/TYZVlGXFwcjh07BpPJhLfeegsNGjS46evZY686rKt8vKmu62cReUuP/dra/vUvICEBWLjQgfBwCQUFAvLzgfx89bGgANcsU5ery4pfU9qywufq47XPi1/rcJR8X+GjVoxGBQEBuCbwi8Pf6QQuXRIQFgZcuaJg6dKK/Q2rtMf+3XffoaCgAJ9++ikOHDiAt99+Gx9++KEnNkXkN6rik0RF3GraqkqbORqF9/d9/nlg2TLgvffy8dBDElwu9ROGJAEul1D0tfpP+Gv5zZcVvs/pxHXLb1xX8evVR6dTXXb4sAHHjol45ZUCt//9PBLs+/btQ6dOnQAArVq1wpEjRzyxGSK/Ulr/NSJC0vxo3Zt3OCNHFo+VdOx4s084Vb/TKdwZqi21AHTu7N6/o0eC/erVqwi+5hqtoijC5XLBeJNhdlEUYLVaKrQtUTRU+L2exLrKh3WVjzfVNWUKUBglhXX16QP06VO8XAv/+5+A1asVdO1qhiga0KePGatXK9i7NxB9+mg3y/v774GRIw1YvVqGzWZA164yhgwJwiefyOja1T3b8MhvPTg4GDk5OUVfy7J801AHAElS2GOvIqyrfFhX+XhTXf/8p/p4+XJxXa1aAa1aqcu0sn27CUuWqDeNlyQLWrXKxZIlIrZvF9GqVflmxdysx25wR6HXa9OmDX744QcAwIEDB9C0aVNPbIaIyOeMHXtjTz0iQnLr3HqPHLH36NEDO3bswKBBg6AoCmbNmuWJzRARUSk8EuwGgwHTp0/3xKqJiOgWPNKKISIi7TDYiYh0hsFORKQzXnHZXiIich8esRMR6QyDnYhIZxjsREQ6w2AnItIZBjsRkc4w2ImIdIbBTkSkMz59z9ODBw9i7ty5SEpK0rqUIk6nEzExMTh79iwKCgowevRo2Gw2rcuCJEl48803ceLECYiiiNmzZyMsLEzrsgAAmZmZ6NevH5YtW4YmTZpoXU6Rp556CtWqqZdFrVevHmbPnq1xRarFixdjy5YtcDqdGDx4MJ599lmtS8L69euxYcMGAEB+fj5+/vln7NixA9WrV9e0LqfTiUmTJuHs2bMwGAyYMWOGV/wfKygowOTJk3H69GkEBwcjNjYWDRs2dNv6fTbYExMTsWnTJgQFBWldSgmbNm2C1WrFu+++i+zsbDz99NNeEexbt24FAKxZswYpKSmYPXu2V9yu0Ol0IjY2FoGBgVqXUkJ+fj4AeNVBAwCkpKQgNTUVq1evRl5eHpYtW6Z1SQCAfv36oV+/fgCAadOm4ZlnntE81AFg27ZtcLlcWLNmDXbs2IEFCxYgPj5e67Kwdu1aWCwWrF27FsePH8eMGTPw0UcfuW39PtuKCQsL84o/0PV69uyJl19+uehrURQ1rKZYZGQkZsyYAQA4d+4catWqpXFFqjlz5mDQoEG46667tC6lhP/973/Iy8vDiBEjMGzYMBw4cEDrkgAAdrsdTZs2xZgxYzBq1Ch0ddctd9zk8OHD+O233zBw4ECtSwEANGrUCJIkQZZlXL16tcwb/lSl3377DZ07dwYANG7cGGlpaW5dv3f8lBXw2GOP4cyZM1qXcYM77rgDgHp7wJdeegmvvPKKxhUVMxqNmDhxIr799lt88MEHWpeD9evXo2bNmujUqROWLFmidTklBAYG4oUXXsCzzz6L33//HdHR0di8ebPmwZCdnY1z584hISEBZ86cwejRo7F582YIgqBpXYUWL16MMWPGaF1GEYvFgrNnz6JXr17Izs5GQkKC1iUBAJo3b46tW7ciMjISBw8eREZGBiRJctuBoM8esXuz9PR0DBs2DE8++SSeeOIJrcspYc6cOfj6668xZcoU5OZqewuzdevWYefOnYiKisLPP/+MiRMn4uLFi5rWVKhRo0bo27cvBEFAo0aNYLVavaI2q9WKiIgImEwmNG7cGGazGVlZWVqXBQC4cuUKjh8/jvbt22tdSpHly5cjIiICX3/9Nf7zn/9g0qRJRW02LT3zzDMIDg7GsGHDsHXrVjzwwANu/XTPYHezS5cuYcSIEZgwYfns7fwAAAX2SURBVAL69++vdTlFNm7ciMWLFwMAgoKCIAiC5m2iVatW4eOPP0ZSUhKaN2+OOXPmIDQ0VNOaCn3++ed4++23AQAZGRm4evWqV9QWHh6O7du3Q1EUZGRkIC8vD1arVeuyAAB79uzBI488onUZJVSvXr1oADwkJAQulwuSJN3iXZ53+PBhhIeHIykpCZGRkahfv75b1++zrRhvlZCQgCtXrmDRokVYtGgRAHWgV+vBwUcffRSTJ0/G0KFD4XK5EBMTA7PZrGlN3qx///6YPHkyBg8eDEEQMGvWLM3bMADQrVs37NmzB/3794eiKIiNjdV8B13oxIkTqFevntZllDB8+HDExMRgyJAhcDqdGDduHCwWi9ZloUGDBnj//fexbNkyVKtWDTNnznTr+nnZXiIinWErhohIZxjsREQ6w2AnItIZBjsRkc4w2ImIdIbBTppLSUnBQw89hPT09KJlc+fOxfr16yu8zjNnzmDAgAHuKO8GkiThhRdewODBg/HHH3+U+dqoqKhyny5+7Ngx7NmzpzIlkp9jsJNXCAgIwOTJk+ELs28vXryI7OxsrF69GiEhIW5f/zfffIPffvvN7esl/6H9GRdEANq3bw9ZlrFq1Sr84x//KFp+5swZvPrqq1i7di0AYMCAAZg3bx42bNiAkydPIjs7G3/88QeGDBmCb775BidOnMCcOXNQq1YtZGVlYdSoUcjKykKXLl0wZswYpKenY8qUKcjPz4fZbMaMGTMgSRJGjx4Nq9WKzp07Izo6umj7mzZtwooVK2AymdCwYUNMnz4dU6ZMwe+//47Y2FhMnz696LUHDx7EzJkzoSgKateujblz5xZ9Lz4+HrVq1cLgwYORlpaGuLg4JCUlYf78+di9ezdkWUbv3r3Rq1cvbNiwAQEBAXjggQfgcDgwf/58iKKI+vXrY/r06fjiiy+wbt06yLKMl156CR06dKiCvxD5EgY7eY24uDg8++yziIiIuK3XBwYG4qOPPsKSJUuwbds2JCQkYN26dfjqq6/w3HPPITc3F++++y4sFguGDh0Km82GhIQEREVFoUuXLti1axfmzp2LcePG4eLFi1i3bh1MJlPR+rOzsxEfH48NGzYgODgYs2bNwqeffoqpU6fi1VdfLRHqADBlyhTMnz8fTZo0wapVq26rBbNx40Z8/PHHqF27NtavX4/atWvj6aefRq1atdCyZUv07NkTn3zyCe68804sWLAAGzZsgNFoRPXq1b3issvknRjs5DVq1KiBmJgYTJo0CW3atCn1Nde2au6//34AQLVq1XDvvfcCUK8HUniRp2bNmhVdJ6Rly5Y4ceIEfvnlFyxevBhLly6FoigICAgAoN5I49pQB4DTp0/j3nvvRXBwMACgbdu2sNvtN71UbmZmZtFNHIYOHXpbP/O8efMwb948XLp0CZ06dSrxvaysLFy4cKHoCqEOhwMdO3ZEWFgYGjVqdFvrJ//EYCev0r17d3z77bfYsGEDJkyYALPZjMzMTEiShJycnBKXar7VpWrT0tKQk5MDs9mMQ4cOYeDAgWjcuDFGjBiBNm3aIC0trWiQ0mC4cbipXr16SEtLQ25uLiwWC3788ccyA/Wuu+7C77//joYNG2LJkiUlXms2m4uuDnn06FEA6l10Nm/ejHnz5kFRFPTu3Ru9e/eGIAiQZRk1atRAnTp1sGjRIlSrVg3JycmwWCxIT08vtV6iQgx28jpvvPEGdu/eDQAIDQ1Fx44d0b9/f4SFhaFBgwa3vZ6QkBCMGzcOWVlZePzxx3Hvvfdi4sSJiIuLQ35+PhwOB954442bvr9mzZoYO3Yshg0bBoPBgLCwMLz22ms3vXzvtGnTEBMTA4PBgNDQUAwfPhwrV64EAPTq1QuvvPIK9uzZgxYtWgAATCYTQkJC8OSTTyIkJAQdO3ZE3bp10aJFC7zzzjto0qQJ3njjDYwcORKKouCOO+7AO++8U2L2EFFpeBEwIiKd4ec5IiKdYbATEekMg52ISGcY7EREOsNgJyLSGQY7EZHOMNiJiHTm/wFnxBsHsxNpNAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Let us now fit kmeans algorithm on our scaled data and Find out the optimum number of clusters to use.\n",
    "\n",
    "# Steps:\n",
    "# 1. Inititalise a dictionary to store the SSE for each k.\n",
    "# 2. Run for a range of Ks and store SSE for each run\n",
    "# 3. Plot the SSE vs K and find the elbow\n",
    "\n",
    "# step 1\n",
    "sse = {} \n",
    "\n",
    "# step 2 - iterate for a range of Ks and fir the scaled data to the algorithm. Use inertia attribute from the clustering object and \n",
    "# store the inertia value for that k \n",
    "\n",
    "for k in range(1, 10):\n",
    "     kmeans = KMeans(n_clusters= k, max_iter=1000, random_state=12).fit(df[features])\n",
    "     sse[k] = kmeans.inertia_\n",
    "\n",
    "# step 3\n",
    "\n",
    "plt.figure()\n",
    "plt.plot(list(sse.keys()), list(sse.values()), 'bx-')\n",
    "plt.xlabel(\"Number of cluster\")\n",
    "plt.ylabel(\"SSE\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "z25ntZ6XyR2a"
   },
   "source": [
    "### 8. Fit the KMeans clustering on the df_scaled data with n_clusters = 3 and store the predictions in the columns 'Labels' (create new column) (3 marks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "cvxudy-wcxfn"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "      <th>Labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.40</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "0              0.49                0.11               0.20   \n",
       "1              0.24                0.22               0.00   \n",
       "2              0.24                0.67               0.20   \n",
       "\n",
       "   Total_visits_online  Total_calls_made  Labels  \n",
       "0                 0.07              0.00       1  \n",
       "1                 0.67              0.90       0  \n",
       "2                 0.20              0.40       1  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Looking at the plot, we can say that for k=3 or 4, elbow point is achieved.\n",
    "# We will fit the kmeans again with k=3 to get the labels\n",
    "\n",
    "kmeans = KMeans(n_clusters=3, max_iter= 1000, random_state=12)\n",
    "\n",
    "kmeans.fit(df_scaled)\n",
    "df_scaled['Labels'] = kmeans.predict(df_scaled) \n",
    "\n",
    "df_scaled.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0F6OrP6a0Sld"
   },
   "source": [
    "### 9. Generate the boxplot for all the features that cleary differentiates among the clusters( 2 mark)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "id": "V_KmaI3DqPf4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# We have generated the labels with kmeans. Let us look at the various features based on the labels. \n",
    "# The labels should be fed to the 'by' attribute \n",
    "\n",
    "\n",
    "df_scaled.boxplot(by = 'Labels', layout = (1,5),figsize=(20,7))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_WkE333w2Bdu"
   },
   "source": [
    "### 10. Implement the Gaussian Mixture Model clustering and generate the boxplot as above (3 marks)\n",
    "\n",
    "- Fit the algorithm with 3 n_components. Predict on the scaled data and store the results as GmmLabels.(2 marks)\n",
    "- Generate the boxplots to understand the clusters similar to the one generated for the kmeans.(1 mark)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "37D1p0rGORjz"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "      <th>Labels</th>\n",
       "      <th>GmmLabels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.40</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "0              0.49                0.11               0.20   \n",
       "1              0.24                0.22               0.00   \n",
       "2              0.24                0.67               0.20   \n",
       "\n",
       "   Total_visits_online  Total_calls_made  Labels  GmmLabels  \n",
       "0                 0.07              0.00       1          1  \n",
       "1                 0.67              0.90       0          0  \n",
       "2                 0.20              0.40       1          1  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.mixture import GaussianMixture\n",
    "\n",
    "gmm = GaussianMixture(n_components= 3, random_state = 12)\n",
    "gmm.fit(df_scaled) \n",
    "df_scaled['GmmLabels'] = gmm.predict(df_scaled)\n",
    "\n",
    "df_scaled.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "id": "yXGbT_A1Pats"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting boxplots with the new GMM based labels\n",
    "\n",
    "new_features = [\"Avg_Credit_Limit\",\t\"Total_Credit_Cards\",\t\"Total_visits_bank\",\t\"Total_visits_online\",\t\"Total_calls_made\",\"GmmLabels\"]\n",
    "\n",
    "df_scaled[new_features].boxplot(by = 'GmmLabels', layout = (1,5),figsize=(20,7))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "abDyu7iL4fWK"
   },
   "source": [
    "### 11. Use the DBSCAN clustering with the eps = 0.5 and other parameters as default on the scaled data. Report the number of clusters formed. (3 marks)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "id": "vEf-sw45Vgg6"
   },
   "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>Avg_Credit_Limit</th>\n",
       "      <th>Total_Credit_Cards</th>\n",
       "      <th>Total_visits_bank</th>\n",
       "      <th>Total_visits_online</th>\n",
       "      <th>Total_calls_made</th>\n",
       "      <th>Labels</th>\n",
       "      <th>GmmLabels</th>\n",
       "      <th>DBSLabels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.24</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.40</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.14</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.56</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.30</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>639</th>\n",
       "      <td>0.49</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>640</th>\n",
       "      <td>0.41</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.20</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>641</th>\n",
       "      <td>0.72</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.60</td>\n",
       "      <td>0.10</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>642</th>\n",
       "      <td>0.86</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.20</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>643</th>\n",
       "      <td>0.83</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.20</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>644 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Avg_Credit_Limit  Total_Credit_Cards  Total_visits_bank  \\\n",
       "0                0.49                0.11               0.20   \n",
       "1                0.24                0.22               0.00   \n",
       "2                0.24                0.67               0.20   \n",
       "3                0.14                0.44               0.20   \n",
       "4                0.49                0.56               0.00   \n",
       "..                ...                 ...                ...   \n",
       "639              0.49                1.00               0.20   \n",
       "640              0.41                1.00               0.20   \n",
       "641              0.72                0.78               0.20   \n",
       "642              0.86                1.00               0.20   \n",
       "643              0.83                0.89               0.00   \n",
       "\n",
       "     Total_visits_online  Total_calls_made  Labels  GmmLabels  DBSLabels  \n",
       "0                   0.07              0.00       1          1          0  \n",
       "1                   0.67              0.90       0          0          1  \n",
       "2                   0.20              0.40       1          1          0  \n",
       "3                   0.07              0.40       0          0          1  \n",
       "4                   0.80              0.30       2          2          2  \n",
       "..                   ...               ...     ...        ...        ...  \n",
       "639                 0.67              0.00       2          2          2  \n",
       "640                 0.87              0.20       2          2          2  \n",
       "641                 0.60              0.10       2          2          2  \n",
       "642                 1.00              0.00       2          2          2  \n",
       "643                 0.80              0.20       2          2          2  \n",
       "\n",
       "[644 rows x 8 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import DBSCAN\n",
    "\n",
    "dbs = DBSCAN(eps = 0.5)\n",
    "\n",
    "df_scaled['DBSLabels'] = dbs.fit_predict(df_scaled)\n",
    "df_scaled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "59bm1gd3XCqp"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x504 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plotting boxplots with the new DBScan based labels\n",
    "\n",
    "new_features = [\"Avg_Credit_Limit\",\t\"Total_Credit_Cards\",\t\"Total_visits_bank\",\t\"Total_visits_online\",\t\"Total_calls_made\",\"DBSLabels\"]\n",
    "\n",
    "df_scaled[new_features].boxplot(by = 'DBSLabels', layout = (1,5),figsize=(20,7))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AsUN-_LuryDS"
   },
   "source": [
    "### 12. Print the cluster sizes and add comments from the different clustering techniques (3 marks)\n",
    "\n",
    "- Print the cluster size (2 marks)\n",
    "- Comments and interpretations (1 mark)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "vELgaXqoXR0H"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cluster sizes in Kmeans:\n",
      "1    373\n",
      "0    222\n",
      "2     49\n",
      "Name: Labels, dtype: int64\n",
      "\n",
      "Cluster sizes in GMM:\n",
      "1    373\n",
      "0    222\n",
      "2     49\n",
      "Name: GmmLabels, dtype: int64\n",
      "\n",
      "Cluster sizes in DBSCAN\n",
      "0    373\n",
      "1    222\n",
      "2     49\n",
      "Name: DBSLabels, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# Use a function that can give the count of all the different values in a column\n",
    "\n",
    "print('Cluster sizes in Kmeans:')\n",
    "print(df_scaled['Labels'].value_counts())\n",
    "print()\n",
    "\n",
    "print('Cluster sizes in GMM:')\n",
    "print(df_scaled['GmmLabels'].value_counts())\n",
    "print()\n",
    "\n",
    "print('Cluster sizes in DBSCAN')\n",
    "print(df_scaled['DBSLabels'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GPw6qHQkvgrE"
   },
   "source": [
    "#### Write your comments and interpretations here\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "if there isn't any problem in my code is here we can see all 3 methods divided the points into the same number in 3 clusters. Maybe the points in each cluster are really simillar and they are diffrent than others and because of that in all three methods we got th same counts. \n"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [
    "MeIShU1-hh0U",
    "KWyGo_wWjkvX",
    "jj7U8RQTlw49",
    "GWbxYcHCmjh6",
    "IkLtnnkcp0cl",
    "skjNQIiMq2Ar",
    "oPSv-4Hat5Wg",
    "z25ntZ6XyR2a",
    "0F6OrP6a0Sld",
    "_WkE333w2Bdu",
    "abDyu7iL4fWK",
    "AsUN-_LuryDS"
   ],
   "name": "Project Unsupervised Learning - Learner's Notebook",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}