Population Income Cluster Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q 3. Section C (Algorithm implementation using packages)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import all necessary libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "from scipy import cluster\n",
    "import sklearn.metrics.pairwise as smp\n",
    "from pyclustering.cluster.kmedians import kmedians\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3(a)\n",
    "# Loading the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AAGE</th>\n",
       "      <th>ACLSWKR</th>\n",
       "      <th>ADTIND</th>\n",
       "      <th>ADTOCC</th>\n",
       "      <th>AHGA</th>\n",
       "      <th>AHRSPAY</th>\n",
       "      <th>AHSCOL</th>\n",
       "      <th>AMARITL</th>\n",
       "      <th>AMJIND</th>\n",
       "      <th>AMJOCC</th>\n",
       "      <th>...</th>\n",
       "      <th>PARENT</th>\n",
       "      <th>PEFNTVTY</th>\n",
       "      <th>PEMNTVTY</th>\n",
       "      <th>PENATVTY</th>\n",
       "      <th>PRCITSHP</th>\n",
       "      <th>SEOTR</th>\n",
       "      <th>VETQVA</th>\n",
       "      <th>VETYN</th>\n",
       "      <th>WKSWORK</th>\n",
       "      <th>YEAR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>73</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>High school graduate</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Widowed</td>\n",
       "      <td>Not in universe or children</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>58</td>\n",
       "      <td>Self-employed-not incorporated</td>\n",
       "      <td>4</td>\n",
       "      <td>34</td>\n",
       "      <td>Some college but no degree</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Divorced</td>\n",
       "      <td>Construction</td>\n",
       "      <td>Precision production craft &amp; repair</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>52</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10th grade</td>\n",
       "      <td>0</td>\n",
       "      <td>High school</td>\n",
       "      <td>Never married</td>\n",
       "      <td>Not in universe or children</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Vietnam</td>\n",
       "      <td>Vietnam</td>\n",
       "      <td>Vietnam</td>\n",
       "      <td>Foreign born- Not a citizen of U S</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Children</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Never married</td>\n",
       "      <td>Not in universe or children</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>...</td>\n",
       "      <td>Both parents present</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Children</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Never married</td>\n",
       "      <td>Not in universe or children</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>...</td>\n",
       "      <td>Both parents present</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 40 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAGE                          ACLSWKR  ADTIND  ADTOCC  \\\n",
       "0    73                  Not in universe       0       0   \n",
       "1    58   Self-employed-not incorporated       4      34   \n",
       "2    18                  Not in universe       0       0   \n",
       "3     9                  Not in universe       0       0   \n",
       "4    10                  Not in universe       0       0   \n",
       "\n",
       "                          AHGA  AHRSPAY            AHSCOL         AMARITL  \\\n",
       "0         High school graduate        0   Not in universe         Widowed   \n",
       "1   Some college but no degree        0   Not in universe        Divorced   \n",
       "2                   10th grade        0       High school   Never married   \n",
       "3                     Children        0   Not in universe   Never married   \n",
       "4                     Children        0   Not in universe   Never married   \n",
       "\n",
       "                         AMJIND                                AMJOCC  ...  \\\n",
       "0   Not in universe or children                       Not in universe  ...   \n",
       "1                  Construction   Precision production craft & repair  ...   \n",
       "2   Not in universe or children                       Not in universe  ...   \n",
       "3   Not in universe or children                       Not in universe  ...   \n",
       "4   Not in universe or children                       Not in universe  ...   \n",
       "\n",
       "                  PARENT        PEFNTVTY        PEMNTVTY        PENATVTY  \\\n",
       "0        Not in universe   United-States   United-States   United-States   \n",
       "1        Not in universe   United-States   United-States   United-States   \n",
       "2        Not in universe         Vietnam         Vietnam         Vietnam   \n",
       "3   Both parents present   United-States   United-States   United-States   \n",
       "4   Both parents present   United-States   United-States   United-States   \n",
       "\n",
       "                               PRCITSHP SEOTR            VETQVA  VETYN  \\\n",
       "0     Native- Born in the United States     0   Not in universe      2   \n",
       "1     Native- Born in the United States     0   Not in universe      2   \n",
       "2   Foreign born- Not a citizen of U S      0   Not in universe      2   \n",
       "3     Native- Born in the United States     0   Not in universe      0   \n",
       "4     Native- Born in the United States     0   Not in universe      0   \n",
       "\n",
       "   WKSWORK YEAR  \n",
       "0        0   95  \n",
       "1       52   94  \n",
       "2        0   95  \n",
       "3        0   94  \n",
       "4        0   94  \n",
       "\n",
       "[5 rows x 40 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Load the dataset into memory\n",
    "df_population=pd.read_csv(\"population.csv\")\n",
    "df_population.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3(b) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# replacing '?' with NaN\n",
    "df_population=df_population.replace(' ?', np.NaN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE            0\n",
       "ACLSWKR         0\n",
       "ADTIND          0\n",
       "ADTOCC          0\n",
       "AHGA            0\n",
       "AHRSPAY         0\n",
       "AHSCOL          0\n",
       "AMARITL         0\n",
       "AMJIND          0\n",
       "AMJOCC          0\n",
       "ARACE           0\n",
       "AREORGN         0\n",
       "ASEX            0\n",
       "AUNMEM          0\n",
       "AUNTYPE         0\n",
       "AWKSTAT         0\n",
       "CAPGAIN         0\n",
       "CAPLOSS         0\n",
       "DIVVAL          0\n",
       "FILESTAT        0\n",
       "GRINREG         0\n",
       "GRINST        708\n",
       "HHDFMX          0\n",
       "HHDREL          0\n",
       "MIGMTR1     99696\n",
       "MIGMTR3     99696\n",
       "MIGMTR4     99696\n",
       "MIGSAME         0\n",
       "MIGSUN      99696\n",
       "NOEMP           0\n",
       "PARENT          0\n",
       "PEFNTVTY     6713\n",
       "PEMNTVTY     6119\n",
       "PENATVTY     3393\n",
       "PRCITSHP        0\n",
       "SEOTR           0\n",
       "VETQVA          0\n",
       "VETYN           0\n",
       "WKSWORK         0\n",
       "YEAR            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "59856.899999999994"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculating 30 percent of the data for removing corresponding columns\n",
    "df_population.shape[0]*.3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_population=df_population.drop(['MIGMTR1','MIGMTR3','MIGMTR4','MIGSUN'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE         int64\n",
       "ACLSWKR     object\n",
       "ADTIND       int64\n",
       "ADTOCC       int64\n",
       "AHGA        object\n",
       "AHRSPAY      int64\n",
       "AHSCOL      object\n",
       "AMARITL     object\n",
       "AMJIND      object\n",
       "AMJOCC      object\n",
       "ARACE       object\n",
       "AREORGN     object\n",
       "ASEX        object\n",
       "AUNMEM      object\n",
       "AUNTYPE     object\n",
       "AWKSTAT     object\n",
       "CAPGAIN      int64\n",
       "CAPLOSS      int64\n",
       "DIVVAL       int64\n",
       "FILESTAT    object\n",
       "GRINREG     object\n",
       "GRINST      object\n",
       "HHDFMX      object\n",
       "HHDREL      object\n",
       "MIGSAME     object\n",
       "NOEMP        int64\n",
       "PARENT      object\n",
       "PEFNTVTY    object\n",
       "PEMNTVTY    object\n",
       "PENATVTY    object\n",
       "PRCITSHP    object\n",
       "SEOTR        int64\n",
       "VETQVA      object\n",
       "VETYN        int64\n",
       "WKSWORK      int64\n",
       "YEAR         int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  Convert numerical data into categorical data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing age column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in AAGE column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for AAGE column:\n",
      "0 90\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in AAGE column:')\n",
    "print(df_population['AAGE'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for AAGE column:')\n",
    "print(df_population['AAGE'].min(), df_population['AAGE'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "34    3489\n",
       "35    3450\n",
       "36    3353\n",
       "31    3351\n",
       "33    3340\n",
       "      ... \n",
       "85     423\n",
       "86     348\n",
       "87     301\n",
       "88     241\n",
       "89     195\n",
       "Name: AAGE, Length: 91, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['AAGE'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11    17777\n",
       "7     16818\n",
       "1     16303\n",
       "8     16166\n",
       "2     15872\n",
       "13    15542\n",
       "3     15334\n",
       "12    14855\n",
       "6     14738\n",
       "9     14651\n",
       "5     13314\n",
       "4     12937\n",
       "10    12377\n",
       "Name: AAGE_bin, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the AAGE column\n",
    "cut_label_AAGE = ['1', '2', '3', '4', '5','6','7','8','9','10','11','12','13']\n",
    "\n",
    "cut_bins = [0, 5, 10, 15, 20, 25, 30, 35, 40, 45,50, 60, 70, 100]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['AAGE_bin'] = pd.cut(df_population['AAGE'], bins=cut_bins, labels=cut_label_AAGE)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['AAGE_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing ADTIND column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in ADTIND column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for ADTIND column:\n",
      "0 51\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in ADTIND column:')\n",
    "print(df_population['ADTIND'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for ADTIND column:')\n",
    "print(df_population['ADTIND'].min(), df_population['ADTIND'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     100684\n",
       "33     17070\n",
       "43      8283\n",
       "4       5984\n",
       "42      4683\n",
       "45      4482\n",
       "29      4209\n",
       "37      4022\n",
       "41      3964\n",
       "32      3596\n",
       "35      3380\n",
       "39      2937\n",
       "34      2765\n",
       "44      2549\n",
       "2       2196\n",
       "11      1764\n",
       "50      1704\n",
       "40      1651\n",
       "47      1644\n",
       "38      1629\n",
       "24      1503\n",
       "12      1350\n",
       "19      1346\n",
       "30      1181\n",
       "31      1178\n",
       "25      1084\n",
       "9        993\n",
       "22       952\n",
       "36       945\n",
       "13       899\n",
       "1        827\n",
       "48       652\n",
       "27       626\n",
       "49       610\n",
       "3        563\n",
       "21       559\n",
       "6        554\n",
       "5        553\n",
       "8        550\n",
       "16       539\n",
       "23       525\n",
       "18       483\n",
       "15       452\n",
       "7        422\n",
       "14       295\n",
       "46       187\n",
       "17       157\n",
       "28       143\n",
       "26       127\n",
       "51        36\n",
       "20        32\n",
       "10         4\n",
       "Name: ADTIND, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['ADTIND'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    100684\n",
       "5     23215\n",
       "7     20147\n",
       "6     19831\n",
       "3     12836\n",
       "2     12646\n",
       "4     10164\n",
       "Name: ADTIND_bin, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the ADTIND column \n",
    "cut_label_ADTIND = ['1', '2', '3', '4', '5','6','7']\n",
    "\n",
    "cut_bins = [-1, 0, 10, 28, 32, 35,42,51]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['ADTIND_bin'] = pd.cut(df_population['ADTIND'], bins=cut_bins, labels=cut_label_ADTIND)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['ADTIND_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing ADTOCC column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in ADTOCC column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for ADTOCC column:\n",
      "0 46\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in ADTOCC column:')\n",
    "print(df_population['ADTOCC'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for ADTOCC column:')\n",
    "print(df_population['ADTOCC'].min(), df_population['ADTOCC'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     100684\n",
       "2       8756\n",
       "26      7887\n",
       "19      5413\n",
       "29      5105\n",
       "36      4145\n",
       "34      4025\n",
       "10      3683\n",
       "16      3445\n",
       "23      3392\n",
       "12      3340\n",
       "33      3325\n",
       "3       3195\n",
       "35      3168\n",
       "38      3003\n",
       "31      2699\n",
       "32      2398\n",
       "37      2234\n",
       "8       2151\n",
       "42      1918\n",
       "30      1897\n",
       "24      1847\n",
       "17      1771\n",
       "28      1661\n",
       "44      1592\n",
       "41      1592\n",
       "43      1382\n",
       "4       1364\n",
       "13      1271\n",
       "18      1083\n",
       "39      1017\n",
       "14       932\n",
       "5        855\n",
       "15       815\n",
       "27       780\n",
       "25       767\n",
       "9        738\n",
       "7        731\n",
       "11       637\n",
       "40       617\n",
       "1        544\n",
       "21       533\n",
       "6        441\n",
       "22       411\n",
       "45       172\n",
       "20        71\n",
       "46        36\n",
       "Name: ADTOCC, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['ADTOCC'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    100684\n",
       "4     18733\n",
       "7     17708\n",
       "5     17330\n",
       "6     15615\n",
       "3     14739\n",
       "2     14714\n",
       "Name: ADTOCC_bin, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the ADTOCC column \n",
    "cut_label_ADTOCC = ['1', '2', '3', '4', '5','6','7']\n",
    "\n",
    "cut_bins = [-1, 0, 5, 15, 25, 30, 35, 46]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['ADTOCC_bin'] = pd.cut(df_population['ADTOCC'], bins=cut_bins, labels=cut_label_ADTOCC)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['ADTOCC_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing AHRSPAY column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in AHRSPAY column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for AHRSPAY column:\n",
      "0 9999\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in AHRSPAY column:')\n",
    "print(df_population['AHRSPAY'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for AHRSPAY column:')\n",
    "print(df_population['AHRSPAY'].min(), df_population['AHRSPAY'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       188219\n",
       "500        734\n",
       "600        546\n",
       "700        534\n",
       "800        507\n",
       "         ...  \n",
       "743          1\n",
       "584          1\n",
       "170          1\n",
       "1265         1\n",
       "2142         1\n",
       "Name: AHRSPAY, Length: 1240, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['AHRSPAY'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    188219\n",
       "2      7157\n",
       "Name: AHRSPAY_bin, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the AHRSPAY column \n",
    "cut_label_AHRSPAY = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0,999]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['AHRSPAY_bin'] = pd.cut(df_population['AHRSPAY'], bins=cut_bins, labels=cut_label_AHRSPAY)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['AHRSPAY_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing CAPGAIN column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in CAPGAIN column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for CAPGAIN column:\n",
      "0 99999\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in CAPGAIN column:')\n",
    "print(df_population['CAPGAIN'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for CAPGAIN column:')\n",
    "print(df_population['CAPGAIN'].min(), df_population['CAPGAIN'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        192144\n",
       "15024       788\n",
       "7688        609\n",
       "7298        582\n",
       "99999       390\n",
       "          ...  \n",
       "2227          1\n",
       "4594          1\n",
       "2387          1\n",
       "1140          1\n",
       "2601          1\n",
       "Name: CAPGAIN, Length: 132, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['CAPGAIN'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    192144\n",
       "2      7379\n",
       "Name: CAPGAIN_bin, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the CAPGAIN column \n",
    "cut_label_CAPGAIN = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 99999]\n",
    "#Creating a new column using the data after binning\n",
    "df_population['CAPGAIN_bin'] = pd.cut(df_population['CAPGAIN'], bins=cut_bins, labels=cut_label_CAPGAIN)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['CAPGAIN_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing CAPLOSS column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in CAPLOSS column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for CAPLOSS column:\n",
      "0 4608\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in CAPLOSS column:')\n",
    "print(df_population['CAPLOSS'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for CAPLOSS column:')\n",
    "print(df_population['CAPLOSS'].min(), df_population['CAPLOSS'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       195617\n",
       "1902       407\n",
       "1977       381\n",
       "1887       364\n",
       "1602       193\n",
       "         ...  \n",
       "974          2\n",
       "155          1\n",
       "2080         1\n",
       "2519         1\n",
       "1911         1\n",
       "Name: CAPLOSS, Length: 113, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['CAPLOSS'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    195617\n",
       "2      3906\n",
       "Name: CAPLOSS_bin, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the CAPLOSS column \n",
    "cut_label_CAPLOSS = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 4608]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['CAPLOSS_bin'] = pd.cut(df_population['CAPLOSS'], bins=cut_bins, labels=cut_label_CAPLOSS)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['CAPLOSS_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing DIVVAL column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in DIVVAL column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for DIVVAL column:\n",
      "0 99999\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in DIVVAL column:')\n",
    "print(df_population['DIVVAL'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for DIVVAL column:')\n",
    "print(df_population['DIVVAL'].min(), df_population['DIVVAL'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        178382\n",
       "100        1148\n",
       "500        1030\n",
       "1000        894\n",
       "200         866\n",
       "          ...  \n",
       "568           1\n",
       "1309          1\n",
       "21100         1\n",
       "5663          1\n",
       "5024          1\n",
       "Name: DIVVAL, Length: 1478, dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['DIVVAL'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    178382\n",
       "2     19155\n",
       "Name: DIVVAL_bin, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the DIVVAL column\n",
    "cut_label_DIVVAL = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 4608]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['DIVVAL_bin'] = pd.cut(df_population['DIVVAL'], bins=cut_bins, labels=cut_label_DIVVAL)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['DIVVAL_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing NOEMP column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in NOEMP column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for NOEMP column:\n",
      "0 6\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in NOEMP column:')\n",
    "print(df_population['NOEMP'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for NOEMP column:')\n",
    "print(df_population['NOEMP'].min(), df_population['NOEMP'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    95983\n",
       "6    36511\n",
       "1    23109\n",
       "4    14379\n",
       "3    13425\n",
       "2    10081\n",
       "5     6035\n",
       "Name: NOEMP, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['NOEMP'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing SEOTR column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in SEOTR column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for SEOTR column:\n",
      "0 2\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in SEOTR column:')\n",
    "print(df_population['SEOTR'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for SEOTR column:')\n",
    "print(df_population['SEOTR'].min(), df_population['SEOTR'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    180672\n",
       "2     16153\n",
       "1      2698\n",
       "Name: SEOTR, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['SEOTR'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing VETYN column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in VETYN column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for VETYN column:\n",
      "0 2\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in VETYN column:')\n",
    "print(df_population['VETYN'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for VETYN column:')\n",
    "print(df_population['VETYN'].min(), df_population['VETYN'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    150130\n",
       "0     47409\n",
       "1      1984\n",
       "Name: VETYN, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['VETYN'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing WKSWORK column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in WKSWORK column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for WKSWORK column:\n",
      "0 52\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in WKSWORK column:')\n",
    "print(df_population['WKSWORK'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for WKSWORK column:')\n",
    "print(df_population['WKSWORK'].min(), df_population['WKSWORK'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     95983\n",
       "52    70314\n",
       "40     2790\n",
       "50     2304\n",
       "26     2268\n",
       "48     1806\n",
       "12     1780\n",
       "30     1378\n",
       "20     1330\n",
       "8      1126\n",
       "36     1108\n",
       "16      945\n",
       "32      883\n",
       "44      845\n",
       "51      819\n",
       "24      767\n",
       "4       757\n",
       "46      708\n",
       "35      704\n",
       "10      694\n",
       "45      669\n",
       "6       646\n",
       "39      602\n",
       "42      573\n",
       "28      568\n",
       "49      509\n",
       "13      496\n",
       "1       464\n",
       "2       458\n",
       "25      457\n",
       "3       417\n",
       "38      380\n",
       "43      374\n",
       "22      370\n",
       "15      353\n",
       "17      331\n",
       "5       309\n",
       "47      278\n",
       "18      272\n",
       "14      257\n",
       "9       239\n",
       "34      230\n",
       "7       152\n",
       "21      135\n",
       "37      123\n",
       "41       88\n",
       "33       81\n",
       "11       78\n",
       "27       76\n",
       "23       67\n",
       "29       63\n",
       "31       51\n",
       "19       48\n",
       "Name: WKSWORK, dtype: int64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['WKSWORK'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    95983\n",
       "4    70314\n",
       "2    17301\n",
       "3    15925\n",
       "Name: WKSWORK_bin, dtype: int64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the WKSWORK column\n",
    "cut_label_WKSWORK = ['1', '2', '3', '4']\n",
    "\n",
    "cut_bins = [-1, 0, 30, 51, 52]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_population['WKSWORK_bin'] = pd.cut(df_population['WKSWORK'], bins=cut_bins, labels=cut_label_WKSWORK)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_population['WKSWORK_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing YEAR column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in YEAR column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for YEAR column:\n",
      "94 95\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in YEAR column:')\n",
    "print(df_population['YEAR'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for YEAR column:')\n",
    "print(df_population['YEAR'].min(), df_population['YEAR'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "94    99827\n",
       "95    99696\n",
       "Name: YEAR, dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population['YEAR'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q3 (c) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imputation, Bucketization, One-Hot Encoding :"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imputation: Using mode for each feature, replace the missing values in each column in population.csv dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# replacing missing values with mode\n",
    "cols = df_population.columns\n",
    "df_population[cols]=df_population[cols].fillna(df_population.mode().iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE           0\n",
       "ACLSWKR        0\n",
       "ADTIND         0\n",
       "ADTOCC         0\n",
       "AHGA           0\n",
       "AHRSPAY        0\n",
       "AHSCOL         0\n",
       "AMARITL        0\n",
       "AMJIND         0\n",
       "AMJOCC         0\n",
       "ARACE          0\n",
       "AREORGN        0\n",
       "ASEX           0\n",
       "AUNMEM         0\n",
       "AUNTYPE        0\n",
       "AWKSTAT        0\n",
       "CAPGAIN        0\n",
       "CAPLOSS        0\n",
       "DIVVAL         0\n",
       "FILESTAT       0\n",
       "GRINREG        0\n",
       "GRINST         0\n",
       "HHDFMX         0\n",
       "HHDREL         0\n",
       "MIGSAME        0\n",
       "NOEMP          0\n",
       "PARENT         0\n",
       "PEFNTVTY       0\n",
       "PEMNTVTY       0\n",
       "PENATVTY       0\n",
       "PRCITSHP       0\n",
       "SEOTR          0\n",
       "VETQVA         0\n",
       "VETYN          0\n",
       "WKSWORK        0\n",
       "YEAR           0\n",
       "AAGE_bin       0\n",
       "ADTIND_bin     0\n",
       "ADTOCC_bin     0\n",
       "AHRSPAY_bin    0\n",
       "CAPGAIN_bin    0\n",
       "CAPLOSS_bin    0\n",
       "DIVVAL_bin     0\n",
       "WKSWORK_bin    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_population.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
       "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
       "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
       "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# selecting columns with object datatypes\n",
    "df_population_obj = df_population.select_dtypes(include=[np.object])\n",
    "df_population_obj.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bucketization:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The categorical values of each of the categorical attributes were studied and these appear to be different ,\n",
    "# hence a logical bucketization appeared not to be possible and therefore was not performed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  One-Hot Encoding:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        ACLSWKR_ Federal government  ACLSWKR_ Local government  \\\n",
      "0                                 0                          0   \n",
      "1                                 0                          0   \n",
      "2                                 0                          0   \n",
      "3                                 0                          0   \n",
      "4                                 0                          0   \n",
      "...                             ...                        ...   \n",
      "199518                            0                          0   \n",
      "199519                            0                          0   \n",
      "199520                            0                          0   \n",
      "199521                            0                          0   \n",
      "199522                            0                          0   \n",
      "\n",
      "        ACLSWKR_ Never worked  ACLSWKR_ Not in universe  ACLSWKR_ Private  \\\n",
      "0                           0                         1                 0   \n",
      "1                           0                         0                 0   \n",
      "2                           0                         1                 0   \n",
      "3                           0                         1                 0   \n",
      "4                           0                         1                 0   \n",
      "...                       ...                       ...               ...   \n",
      "199518                      0                         1                 0   \n",
      "199519                      0                         0                 0   \n",
      "199520                      0                         1                 0   \n",
      "199521                      0                         1                 0   \n",
      "199522                      0                         0                 1   \n",
      "\n",
      "        ACLSWKR_ Self-employed-incorporated  \\\n",
      "0                                         0   \n",
      "1                                         0   \n",
      "2                                         0   \n",
      "3                                         0   \n",
      "4                                         0   \n",
      "...                                     ...   \n",
      "199518                                    0   \n",
      "199519                                    1   \n",
      "199520                                    0   \n",
      "199521                                    0   \n",
      "199522                                    0   \n",
      "\n",
      "        ACLSWKR_ Self-employed-not incorporated  ACLSWKR_ State government  \\\n",
      "0                                             0                          0   \n",
      "1                                             1                          0   \n",
      "2                                             0                          0   \n",
      "3                                             0                          0   \n",
      "4                                             0                          0   \n",
      "...                                         ...                        ...   \n",
      "199518                                        0                          0   \n",
      "199519                                        0                          0   \n",
      "199520                                        0                          0   \n",
      "199521                                        0                          0   \n",
      "199522                                        0                          0   \n",
      "\n",
      "        ACLSWKR_ Without pay  AHGA_ 10th grade  ...  PENATVTY_ Vietnam  \\\n",
      "0                          0                 0  ...                  0   \n",
      "1                          0                 0  ...                  0   \n",
      "2                          0                 1  ...                  1   \n",
      "3                          0                 0  ...                  0   \n",
      "4                          0                 0  ...                  0   \n",
      "...                      ...               ...  ...                ...   \n",
      "199518                     0                 0  ...                  0   \n",
      "199519                     0                 0  ...                  0   \n",
      "199520                     0                 0  ...                  0   \n",
      "199521                     0                 1  ...                  0   \n",
      "199522                     0                 0  ...                  0   \n",
      "\n",
      "        PENATVTY_ Yugoslavia  PRCITSHP_ Foreign born- Not a citizen of U S   \\\n",
      "0                          0                                              0   \n",
      "1                          0                                              0   \n",
      "2                          0                                              1   \n",
      "3                          0                                              0   \n",
      "4                          0                                              0   \n",
      "...                      ...                                            ...   \n",
      "199518                     0                                              0   \n",
      "199519                     0                                              0   \n",
      "199520                     0                                              0   \n",
      "199521                     0                                              0   \n",
      "199522                     0                                              1   \n",
      "\n",
      "        PRCITSHP_ Foreign born- U S citizen by naturalization  \\\n",
      "0                                                       0       \n",
      "1                                                       0       \n",
      "2                                                       0       \n",
      "3                                                       0       \n",
      "4                                                       0       \n",
      "...                                                   ...       \n",
      "199518                                                  0       \n",
      "199519                                                  0       \n",
      "199520                                                  1       \n",
      "199521                                                  0       \n",
      "199522                                                  0       \n",
      "\n",
      "        PRCITSHP_ Native- Born abroad of American Parent(s)  \\\n",
      "0                                                       0     \n",
      "1                                                       0     \n",
      "2                                                       0     \n",
      "3                                                       0     \n",
      "4                                                       0     \n",
      "...                                                   ...     \n",
      "199518                                                  0     \n",
      "199519                                                  0     \n",
      "199520                                                  0     \n",
      "199521                                                  0     \n",
      "199522                                                  0     \n",
      "\n",
      "        PRCITSHP_ Native- Born in Puerto Rico or U S Outlying  \\\n",
      "0                                                       0       \n",
      "1                                                       0       \n",
      "2                                                       0       \n",
      "3                                                       0       \n",
      "4                                                       0       \n",
      "...                                                   ...       \n",
      "199518                                                  0       \n",
      "199519                                                  0       \n",
      "199520                                                  0       \n",
      "199521                                                  0       \n",
      "199522                                                  0       \n",
      "\n",
      "        PRCITSHP_ Native- Born in the United States  VETQVA_ No  \\\n",
      "0                                                 1           0   \n",
      "1                                                 1           0   \n",
      "2                                                 0           0   \n",
      "3                                                 1           0   \n",
      "4                                                 1           0   \n",
      "...                                             ...         ...   \n",
      "199518                                            1           0   \n",
      "199519                                            1           0   \n",
      "199520                                            0           0   \n",
      "199521                                            1           0   \n",
      "199522                                            0           0   \n",
      "\n",
      "        VETQVA_ Not in universe  VETQVA_ Yes  \n",
      "0                             1            0  \n",
      "1                             1            0  \n",
      "2                             1            0  \n",
      "3                             1            0  \n",
      "4                             1            0  \n",
      "...                         ...          ...  \n",
      "199518                        1            0  \n",
      "199519                        1            0  \n",
      "199520                        1            0  \n",
      "199521                        1            0  \n",
      "199522                        1            0  \n",
      "\n",
      "[199523 rows x 359 columns]\n"
     ]
    }
   ],
   "source": [
    "# one hot encode the required features\n",
    "df_population_one_hot_encoded_data = pd.get_dummies(df_population[['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
    "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
    "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
    "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA']], columns = ['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
    "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
    "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
    "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA'])\n",
    "print(df_population_one_hot_encoded_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# selecting columns with integer datatypes\n",
    "df_population_int =df_population[['AAGE_bin', 'ADTOCC_bin', 'ADTIND_bin',\n",
    "       'AHRSPAY_bin', 'CAPGAIN_bin', 'CAPLOSS_bin', 'DIVVAL_bin', 'NOEMP',\n",
    "       'SEOTR', 'VETYN', 'WKSWORK_bin', 'YEAR']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create concatenated dataframe\n",
    "data_population = pd.concat([df_population_int, df_population_one_hot_encoded_data], axis=1) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q3 (d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_population_1 = data_population.copy()\n",
    "data_population_1['Label'] = -1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performing Principal Component Analysis "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=2)\n",
    "data_population_pca = pca.fit_transform(data_population)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finding the average within-cluster distance for the number of clusters (k) in the range 10 to 25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Create empty lists to store cluster size and average within-cluster distances \n",
    "ks = []\n",
    "avg_dist = []\n",
    "\n",
    "# The kmedians function requires the initial value of medians to be given to it as an input.\n",
    "# However, no such value are given in the data, hence, we have used kmeans where k is ranging from 10 to 24 to find out\n",
    "# the set of centroids per value of k. These centroid values have been given as input to the kmedians function assuming\n",
    "# the centroids in the kmeans are the most central points in the clusters formed.\n",
    "\n",
    "for k in range(10,25):\n",
    "    model = KMeans(n_clusters=k, random_state=0).fit(data_population_pca)\n",
    "    centers = np.array(model.cluster_centers_)\n",
    "    ks.append(k)\n",
    "    # Create instance of K-Medians algorithm.\n",
    "    kmedians_instance = kmedians(data_population_pca, initial_medians = centers)\n",
    "    # Run cluster analysis and obtain results.\n",
    "    kmedians_instance.process()\n",
    "    clusters = kmedians_instance.get_clusters()\n",
    "    medians = kmedians_instance.get_medians()\n",
    "    for i in range(len(clusters)):\n",
    "        data_population_1.iloc[data_population_1.index.isin(clusters[i]),-1]=i\n",
    "    new_df_population = pd.DataFrame(data_population_pca, columns=['PC1','PC2'])\n",
    "    new_df_population['Label'] = data_population_1['Label']\n",
    "    dists=[]\n",
    "    for i, (mx, my) in enumerate(medians):\n",
    "        a= pd.DataFrame()\n",
    "        a=new_df_population[new_df_population['Label'] == i].iloc[:, :-1]\n",
    "        # To find out average within-cluster distance we have used dist = euclidean distance between (PC1, PC2) and (median1, median2)\n",
    "        a['dist'] = np.sqrt((a['PC1']-medians[i][0])**2+(a['PC2']-medians[i][1])**2)\n",
    "        dists.append(np.mean(a['dist']))\n",
    "    avg_dist.append(np.mean(dists))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting the average within-cluster distance vs. a number of clusters graph.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(ks, avg_dist, 'bx-') \n",
    "plt.xlabel('Values of K') \n",
    "plt.ylabel('average within-cluster distance') \n",
    "plt.title('the average within-cluster distance vs. a number of clusters graph. ') \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Choosing the best value of k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From the above plot , the best value of k appears to be at k=24.However, as a comparision has to be made between the\n",
    "# clusters formed between population.csv and  more_than_50k.csv, so we tried to choose that particular value of k which\n",
    "# was optimally good and at the same time common across both the datasets.\n",
    "\n",
    "#For more_than_50k.csv the best value of k came at k=22, so we decided to use k=22 for both the datasets so that direct\n",
    "#comparision can be made between the clusters of the two datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Q3 (e)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Handling more_than_50k data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# All the steps adopted on population.csv as per parts a, b, c and d have been duplicated in part e for more_than_50k.csv "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>AAGE</th>\n",
       "      <th>ACLSWKR</th>\n",
       "      <th>ADTIND</th>\n",
       "      <th>ADTOCC</th>\n",
       "      <th>AHGA</th>\n",
       "      <th>AHRSPAY</th>\n",
       "      <th>AHSCOL</th>\n",
       "      <th>AMARITL</th>\n",
       "      <th>AMJIND</th>\n",
       "      <th>AMJOCC</th>\n",
       "      <th>...</th>\n",
       "      <th>PARENT</th>\n",
       "      <th>PEFNTVTY</th>\n",
       "      <th>PEMNTVTY</th>\n",
       "      <th>PENATVTY</th>\n",
       "      <th>PRCITSHP</th>\n",
       "      <th>SEOTR</th>\n",
       "      <th>VETQVA</th>\n",
       "      <th>VETYN</th>\n",
       "      <th>WKSWORK</th>\n",
       "      <th>YEAR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>57</td>\n",
       "      <td>Self-employed-incorporated</td>\n",
       "      <td>11</td>\n",
       "      <td>2</td>\n",
       "      <td>High school graduate</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Married-civilian spouse present</td>\n",
       "      <td>Manufacturing-durable goods</td>\n",
       "      <td>Executive admin and managerial</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>52</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>44</td>\n",
       "      <td>Private</td>\n",
       "      <td>3</td>\n",
       "      <td>38</td>\n",
       "      <td>High school graduate</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Married-civilian spouse present</td>\n",
       "      <td>Mining</td>\n",
       "      <td>Transportation and material moving</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>54</td>\n",
       "      <td>Private</td>\n",
       "      <td>19</td>\n",
       "      <td>26</td>\n",
       "      <td>High school graduate</td>\n",
       "      <td>1550</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Married-civilian spouse present</td>\n",
       "      <td>Manufacturing-nondurable goods</td>\n",
       "      <td>Adm support including clerical</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>52</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>Private</td>\n",
       "      <td>33</td>\n",
       "      <td>2</td>\n",
       "      <td>9th grade</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Never married</td>\n",
       "      <td>Retail trade</td>\n",
       "      <td>Executive admin and managerial</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>1</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>52</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26</td>\n",
       "      <td>Private</td>\n",
       "      <td>11</td>\n",
       "      <td>5</td>\n",
       "      <td>Bachelors degree(BA AB BS)</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>Married-civilian spouse present</td>\n",
       "      <td>Manufacturing-durable goods</td>\n",
       "      <td>Professional specialty</td>\n",
       "      <td>...</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>United-States</td>\n",
       "      <td>Native- Born in the United States</td>\n",
       "      <td>0</td>\n",
       "      <td>Not in universe</td>\n",
       "      <td>2</td>\n",
       "      <td>52</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 40 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAGE                      ACLSWKR  ADTIND  ADTOCC  \\\n",
       "0    57   Self-employed-incorporated      11       2   \n",
       "1    44                      Private       3      38   \n",
       "2    54                      Private      19      26   \n",
       "3    25                      Private      33       2   \n",
       "4    26                      Private      11       5   \n",
       "\n",
       "                          AHGA  AHRSPAY            AHSCOL  \\\n",
       "0         High school graduate        0   Not in universe   \n",
       "1         High school graduate        0   Not in universe   \n",
       "2         High school graduate     1550   Not in universe   \n",
       "3                    9th grade        0   Not in universe   \n",
       "4   Bachelors degree(BA AB BS)        0   Not in universe   \n",
       "\n",
       "                            AMARITL                           AMJIND  \\\n",
       "0   Married-civilian spouse present      Manufacturing-durable goods   \n",
       "1   Married-civilian spouse present                           Mining   \n",
       "2   Married-civilian spouse present   Manufacturing-nondurable goods   \n",
       "3                     Never married                     Retail trade   \n",
       "4   Married-civilian spouse present      Manufacturing-durable goods   \n",
       "\n",
       "                                AMJOCC  ...            PARENT        PEFNTVTY  \\\n",
       "0       Executive admin and managerial  ...   Not in universe   United-States   \n",
       "1   Transportation and material moving  ...   Not in universe   United-States   \n",
       "2       Adm support including clerical  ...   Not in universe   United-States   \n",
       "3       Executive admin and managerial  ...   Not in universe   United-States   \n",
       "4               Professional specialty  ...   Not in universe   United-States   \n",
       "\n",
       "         PEMNTVTY        PENATVTY                            PRCITSHP SEOTR  \\\n",
       "0   United-States   United-States   Native- Born in the United States     0   \n",
       "1   United-States   United-States   Native- Born in the United States     0   \n",
       "2   United-States   United-States   Native- Born in the United States     0   \n",
       "3   United-States   United-States   Native- Born in the United States     1   \n",
       "4   United-States   United-States   Native- Born in the United States     0   \n",
       "\n",
       "             VETQVA  VETYN  WKSWORK YEAR  \n",
       "0   Not in universe      2       52   95  \n",
       "1                No      1       48   94  \n",
       "2   Not in universe      2       52   94  \n",
       "3   Not in universe      2       52   95  \n",
       "4   Not in universe      2       52   95  \n",
       "\n",
       "[5 rows x 40 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k=pd.read_csv(\"more_than_50k.csv\")\n",
    "df_more_than_50k.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_more_than_50k=df_more_than_50k.replace(' ?', np.NaN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE           0\n",
       "ACLSWKR        0\n",
       "ADTIND         0\n",
       "ADTOCC         0\n",
       "AHGA           0\n",
       "AHRSPAY        0\n",
       "AHSCOL         0\n",
       "AMARITL        0\n",
       "AMJIND         0\n",
       "AMJOCC         0\n",
       "ARACE          0\n",
       "AREORGN        0\n",
       "ASEX           0\n",
       "AUNMEM         0\n",
       "AUNTYPE        0\n",
       "AWKSTAT        0\n",
       "CAPGAIN        0\n",
       "CAPLOSS        0\n",
       "DIVVAL         0\n",
       "FILESTAT       0\n",
       "GRINREG        0\n",
       "GRINST        14\n",
       "HHDFMX         0\n",
       "HHDREL         0\n",
       "MIGMTR1     1906\n",
       "MIGMTR3     1906\n",
       "MIGMTR4     1906\n",
       "MIGSAME        0\n",
       "MIGSUN      1906\n",
       "NOEMP          0\n",
       "PARENT         0\n",
       "PEFNTVTY     162\n",
       "PEMNTVTY     134\n",
       "PENATVTY      93\n",
       "PRCITSHP       0\n",
       "SEOTR          0\n",
       "VETQVA         0\n",
       "VETYN          0\n",
       "WKSWORK        0\n",
       "YEAR           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1074.0"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k.shape[0]*.3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_more_than_50k=df_more_than_50k.drop(['MIGMTR1','MIGMTR3','MIGMTR4','MIGSUN'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['AAGE', 'ACLSWKR', 'ADTIND', 'ADTOCC', 'AHGA', 'AHRSPAY', 'AHSCOL',\n",
       "       'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE', 'AREORGN', 'ASEX', 'AUNMEM',\n",
       "       'AUNTYPE', 'AWKSTAT', 'CAPGAIN', 'CAPLOSS', 'DIVVAL', 'FILESTAT',\n",
       "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'NOEMP', 'PARENT',\n",
       "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'SEOTR', 'VETQVA',\n",
       "       'VETYN', 'WKSWORK', 'YEAR'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE         int64\n",
       "ACLSWKR     object\n",
       "ADTIND       int64\n",
       "ADTOCC       int64\n",
       "AHGA        object\n",
       "AHRSPAY      int64\n",
       "AHSCOL      object\n",
       "AMARITL     object\n",
       "AMJIND      object\n",
       "AMJOCC      object\n",
       "ARACE       object\n",
       "AREORGN     object\n",
       "ASEX        object\n",
       "AUNMEM      object\n",
       "AUNTYPE     object\n",
       "AWKSTAT     object\n",
       "CAPGAIN      int64\n",
       "CAPLOSS      int64\n",
       "DIVVAL       int64\n",
       "FILESTAT    object\n",
       "GRINREG     object\n",
       "GRINST      object\n",
       "HHDFMX      object\n",
       "HHDREL      object\n",
       "MIGSAME     object\n",
       "NOEMP        int64\n",
       "PARENT      object\n",
       "PEFNTVTY    object\n",
       "PEMNTVTY    object\n",
       "PENATVTY    object\n",
       "PRCITSHP    object\n",
       "SEOTR        int64\n",
       "VETQVA      object\n",
       "VETYN        int64\n",
       "WKSWORK      int64\n",
       "YEAR         int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  Convert numerical data into categorical data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing age column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in AAGE column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for AAGE column:\n",
      "18 90\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in AAGE column:')\n",
    "print(df_more_than_50k['AAGE'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for AAGE column:')\n",
    "print(df_more_than_50k['AAGE'].min(), df_more_than_50k['AAGE'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "46    154\n",
       "47    140\n",
       "39    132\n",
       "37    124\n",
       "45    123\n",
       "     ... \n",
       "20      1\n",
       "89      1\n",
       "88      1\n",
       "83      1\n",
       "18      1\n",
       "Name: AAGE, Length: 73, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['AAGE'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6    700\n",
       "5    626\n",
       "3    589\n",
       "4    571\n",
       "2    439\n",
       "7    277\n",
       "1    241\n",
       "8    136\n",
       "Name: AAGE_bin, dtype: int64"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the AAGE column \n",
    "\n",
    "cut_label_AAGE = ['1', '2', '3', '4', '5','6','7','8']\n",
    "\n",
    "cut_bins = [18, 30, 35, 40, 45,50, 60, 70, 90]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['AAGE_bin'] = pd.cut(df_more_than_50k['AAGE'], bins=cut_bins, labels=cut_label_AAGE)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['AAGE_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing ADTIND column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in ADTIND column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for ADTIND column:\n",
      "0 51\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in ADTIND column:')\n",
    "print(df_more_than_50k['ADTIND'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for ADTIND column:')\n",
    "print(df_more_than_50k['ADTIND'].min(), df_more_than_50k['ADTIND'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45    321\n",
       "0     275\n",
       "43    267\n",
       "33    210\n",
       "32    170\n",
       "4     168\n",
       "42    163\n",
       "35    160\n",
       "41    158\n",
       "34    157\n",
       "37    127\n",
       "29    125\n",
       "47     92\n",
       "12     85\n",
       "50     83\n",
       "11     83\n",
       "25     83\n",
       "30     78\n",
       "31     72\n",
       "13     62\n",
       "16     45\n",
       "3      44\n",
       "49     44\n",
       "24     41\n",
       "38     36\n",
       "9      35\n",
       "44     32\n",
       "40     31\n",
       "2      31\n",
       "39     30\n",
       "19     26\n",
       "23     26\n",
       "15     26\n",
       "48     23\n",
       "8      22\n",
       "14     22\n",
       "26     16\n",
       "27     15\n",
       "5      14\n",
       "46     13\n",
       "18     11\n",
       "1      10\n",
       "7      10\n",
       "21      9\n",
       "22      7\n",
       "6       7\n",
       "20      5\n",
       "17      4\n",
       "51      3\n",
       "28      2\n",
       "36      1\n",
       "Name: ADTIND, dtype: int64"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['ADTIND'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7    878\n",
       "3    568\n",
       "6    546\n",
       "5    527\n",
       "4    445\n",
       "2    341\n",
       "1    275\n",
       "Name: ADTIND_bin, dtype: int64"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the ADTIND column \n",
    "cut_label_ADTIND = ['1', '2', '3', '4', '5','6','7']\n",
    "\n",
    "cut_bins = [-1, 0, 10, 28, 32, 35,42,51]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['ADTIND_bin'] = pd.cut(df_more_than_50k['ADTIND'], bins=cut_bins, labels=cut_label_ADTIND)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['ADTIND_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing ADTOCC column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in ADTOCC column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for ADTOCC column:\n",
      "0 46\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in ADTOCC column:')\n",
    "print(df_more_than_50k['ADTOCC'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for ADTOCC column:')\n",
    "print(df_more_than_50k['ADTOCC'].min(), df_more_than_50k['ADTOCC'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2     820\n",
       "0     275\n",
       "4     189\n",
       "3     185\n",
       "16    174\n",
       "11    137\n",
       "12    136\n",
       "7     134\n",
       "17    130\n",
       "10    120\n",
       "8     115\n",
       "34    104\n",
       "33    101\n",
       "18     91\n",
       "35     82\n",
       "5      79\n",
       "28     73\n",
       "9      60\n",
       "36     52\n",
       "26     50\n",
       "15     47\n",
       "6      46\n",
       "38     40\n",
       "14     38\n",
       "19     35\n",
       "1      31\n",
       "43     29\n",
       "21     27\n",
       "39     25\n",
       "37     22\n",
       "13     17\n",
       "23     16\n",
       "30     11\n",
       "25     11\n",
       "32     10\n",
       "24     10\n",
       "45      8\n",
       "29      8\n",
       "44      8\n",
       "31      7\n",
       "22      6\n",
       "41      6\n",
       "40      5\n",
       "42      5\n",
       "46      3\n",
       "20      2\n",
       "Name: ADTOCC, dtype: int64"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['ADTOCC'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    1304\n",
       "3     850\n",
       "4     502\n",
       "6     304\n",
       "1     275\n",
       "7     203\n",
       "5     142\n",
       "Name: ADTOCC_bin, dtype: int64"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the ADTOCC column \n",
    "cut_label_ADTOCC = ['1', '2', '3', '4', '5','6','7']\n",
    "\n",
    "cut_bins = [-1, 0, 5, 15, 25, 30, 35, 46]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['ADTOCC_bin'] = pd.cut(df_more_than_50k['ADTOCC'], bins=cut_bins, labels=cut_label_ADTOCC)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['ADTOCC_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing AHRSPAY column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in AHRSPAY column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for AHRSPAY column:\n",
      "0 8000\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in AHRSPAY column:')\n",
    "print(df_more_than_50k['AHRSPAY'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for AHRSPAY column:')\n",
    "print(df_more_than_50k['AHRSPAY'].min(), df_more_than_50k['AHRSPAY'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       3431\n",
       "2100       9\n",
       "2000       8\n",
       "1700       5\n",
       "2200       5\n",
       "        ... \n",
       "525        1\n",
       "585        1\n",
       "625        1\n",
       "2700       1\n",
       "2010       1\n",
       "Name: AHRSPAY, Length: 85, dtype: int64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['AHRSPAY'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    3431\n",
       "2      32\n",
       "Name: AHRSPAY_bin, dtype: int64"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the AHRSPAY column \n",
    "cut_label_AHRSPAY = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0,999]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['AHRSPAY_bin'] = pd.cut(df_more_than_50k['AHRSPAY'], bins=cut_bins, labels=cut_label_AHRSPAY)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['AHRSPAY_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing CAPGAIN column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in CAPGAIN column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for CAPGAIN column:\n",
      "0 99999\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in CAPGAIN column:')\n",
    "print(df_more_than_50k['CAPGAIN'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for CAPGAIN column:')\n",
    "print(df_more_than_50k['CAPGAIN'].min(), df_more_than_50k['CAPGAIN'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        2902\n",
       "15024     193\n",
       "7688       97\n",
       "99999      93\n",
       "7298       54\n",
       "27828      32\n",
       "14084      26\n",
       "10520      24\n",
       "20051      21\n",
       "4787       20\n",
       "13550      20\n",
       "10605      14\n",
       "5178       10\n",
       "9386       10\n",
       "3103        8\n",
       "4934        7\n",
       "7430        5\n",
       "18481       5\n",
       "6514        5\n",
       "6097        4\n",
       "25124       4\n",
       "25236       4\n",
       "11678       3\n",
       "15831       2\n",
       "4687        2\n",
       "15020       2\n",
       "7896        2\n",
       "7262        2\n",
       "4386        1\n",
       "8614        1\n",
       "9562        1\n",
       "2653        1\n",
       "3273        1\n",
       "3456        1\n",
       "5556        1\n",
       "6612        1\n",
       "4064        1\n",
       "Name: CAPGAIN, dtype: int64"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['CAPGAIN'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    2902\n",
       "2     678\n",
       "Name: CAPGAIN_bin, dtype: int64"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the CAPGAIN column \n",
    "cut_label_CAPGAIN = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 99999]\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['CAPGAIN_bin'] = pd.cut(df_more_than_50k['CAPGAIN'], bins=cut_bins, labels=cut_label_CAPGAIN)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['CAPGAIN_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing CAPLOSS column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in CAPLOSS column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for CAPLOSS column:\n",
      "0 3004\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in CAPLOSS column:')\n",
    "print(df_more_than_50k['CAPLOSS'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for CAPLOSS column:')\n",
    "print(df_more_than_50k['CAPLOSS'].min(), df_more_than_50k['CAPLOSS'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       3243\n",
       "1902      89\n",
       "1977      87\n",
       "2415      32\n",
       "1887      28\n",
       "2258      19\n",
       "2824      12\n",
       "1564      10\n",
       "2559       9\n",
       "2444       6\n",
       "2174       5\n",
       "2377       4\n",
       "2472       3\n",
       "2392       3\n",
       "323        3\n",
       "2282       3\n",
       "3004       3\n",
       "2246       2\n",
       "1825       2\n",
       "653        2\n",
       "1848       2\n",
       "2547       2\n",
       "1408       2\n",
       "1092       1\n",
       "2231       1\n",
       "1669       1\n",
       "1380       1\n",
       "1740       1\n",
       "1876       1\n",
       "1485       1\n",
       "625        1\n",
       "2090       1\n",
       "Name: CAPLOSS, dtype: int64"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['CAPLOSS'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    3243\n",
       "2     337\n",
       "Name: CAPLOSS_bin, dtype: int64"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the CAPLOSS column \n",
    "cut_label_CAPLOSS = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 4608]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['CAPLOSS_bin'] = pd.cut(df_more_than_50k['CAPLOSS'], bins=cut_bins, labels=cut_label_CAPLOSS)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['CAPLOSS_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing DIVVAL column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in DIVVAL column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for DIVVAL column:\n",
      "0 99999\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in DIVVAL column:')\n",
    "print(df_more_than_50k['DIVVAL'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for DIVVAL column:')\n",
    "print(df_more_than_50k['DIVVAL'].min(), df_more_than_50k['DIVVAL'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        2023\n",
       "1000      104\n",
       "100        81\n",
       "500        77\n",
       "200        65\n",
       "         ... \n",
       "2022        1\n",
       "7616        1\n",
       "6062        1\n",
       "12173       1\n",
       "1700        1\n",
       "Name: DIVVAL, Length: 302, dtype: int64"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['DIVVAL'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    2023\n",
       "2    1292\n",
       "Name: DIVVAL_bin, dtype: int64"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the DIVVAL column \n",
    "cut_label_DIVVAL = ['1', '2']\n",
    "\n",
    "cut_bins = [-1, 0, 4608]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['DIVVAL_bin'] = pd.cut(df_more_than_50k['DIVVAL'], bins=cut_bins, labels=cut_label_DIVVAL)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['DIVVAL_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing NOEMP column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in NOEMP column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for NOEMP column:\n",
      "0 6\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in NOEMP column:')\n",
    "print(df_more_than_50k['NOEMP'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for NOEMP column:')\n",
    "print(df_more_than_50k['NOEMP'].min(), df_more_than_50k['NOEMP'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6    1601\n",
       "1     592\n",
       "4     429\n",
       "3     353\n",
       "2     232\n",
       "5     196\n",
       "0     177\n",
       "Name: NOEMP, dtype: int64"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['NOEMP'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing SEOTR column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in SEOTR column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for SEOTR column:\n",
      "0 2\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in SEOTR column:')\n",
    "print(df_more_than_50k['SEOTR'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for SEOTR column:')\n",
    "print(df_more_than_50k['SEOTR'].min(), df_more_than_50k['SEOTR'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    3028\n",
       "2     382\n",
       "1     170\n",
       "Name: SEOTR, dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['SEOTR'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing VETYN column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in VETYN column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for VETYN column:\n",
      "1 2\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in VETYN column:')\n",
    "print(df_more_than_50k['VETYN'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for VETYN column:')\n",
    "print(df_more_than_50k['VETYN'].min(), df_more_than_50k['VETYN'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    3517\n",
       "1      63\n",
       "Name: VETYN, dtype: int64"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['VETYN'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing WKSWORK column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in WKSWORK column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for WKSWORK column:\n",
      "0 52\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in WKSWORK column:')\n",
    "print(df_more_than_50k['WKSWORK'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for WKSWORK column:')\n",
    "print(df_more_than_50k['WKSWORK'].min(), df_more_than_50k['WKSWORK'].max())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "52    2977\n",
       "0      177\n",
       "50      68\n",
       "48      51\n",
       "40      50\n",
       "26      28\n",
       "51      25\n",
       "46      23\n",
       "49      16\n",
       "30      13\n",
       "32      13\n",
       "36      13\n",
       "44      10\n",
       "45       9\n",
       "8        8\n",
       "28       7\n",
       "39       7\n",
       "24       7\n",
       "20       7\n",
       "47       6\n",
       "12       6\n",
       "35       5\n",
       "1        4\n",
       "16       4\n",
       "42       4\n",
       "10       4\n",
       "13       4\n",
       "41       3\n",
       "29       3\n",
       "38       3\n",
       "34       3\n",
       "37       2\n",
       "21       2\n",
       "25       2\n",
       "3        2\n",
       "7        2\n",
       "43       2\n",
       "17       1\n",
       "33       1\n",
       "27       1\n",
       "22       1\n",
       "18       1\n",
       "14       1\n",
       "11       1\n",
       "15       1\n",
       "4        1\n",
       "2        1\n",
       "Name: WKSWORK, dtype: int64"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['WKSWORK'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4    2977\n",
       "3     314\n",
       "1     177\n",
       "2     112\n",
       "Name: WKSWORK_bin, dtype: int64"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Discretizing the WKSWORK column \n",
    "cut_label_WKSWORK = ['1', '2', '3', '4']\n",
    "\n",
    "cut_bins = [-1, 0, 30, 51, 52]\n",
    "\n",
    "#Creating a new column using the data after binning\n",
    "df_more_than_50k['WKSWORK_bin'] = pd.cut(df_more_than_50k['WKSWORK'], bins=cut_bins, labels=cut_label_WKSWORK)\n",
    "\n",
    "# Finding counts of unique values\n",
    "df_more_than_50k['WKSWORK_bin'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing YEAR column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Null Values in YEAR column:\n",
      "0\n",
      "\n",
      "Minimum and Maximum values for YEAR column:\n",
      "94 95\n"
     ]
    }
   ],
   "source": [
    "#Display total null values\n",
    "print('Total Null Values in YEAR column:')\n",
    "print(df_more_than_50k['YEAR'].isna().sum())\n",
    "\n",
    "# Display minimum and maximum values\n",
    "print('\\nMinimum and Maximum values for YEAR column:')\n",
    "print(df_more_than_50k['YEAR'].min(), df_more_than_50k['YEAR'].max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "95    1906\n",
       "94    1674\n",
       "Name: YEAR, dtype: int64"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k['YEAR'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imputation, Bucketization, One-Hot Encoding :"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imputation: Using mode for each feature, replace the missing values in each column in both the datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = df_more_than_50k.columns\n",
    "df_more_than_50k[cols]=df_more_than_50k[cols].fillna(df_more_than_50k.mode().iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AAGE           0\n",
       "ACLSWKR        0\n",
       "ADTIND         0\n",
       "ADTOCC         0\n",
       "AHGA           0\n",
       "AHRSPAY        0\n",
       "AHSCOL         0\n",
       "AMARITL        0\n",
       "AMJIND         0\n",
       "AMJOCC         0\n",
       "ARACE          0\n",
       "AREORGN        0\n",
       "ASEX           0\n",
       "AUNMEM         0\n",
       "AUNTYPE        0\n",
       "AWKSTAT        0\n",
       "CAPGAIN        0\n",
       "CAPLOSS        0\n",
       "DIVVAL         0\n",
       "FILESTAT       0\n",
       "GRINREG        0\n",
       "GRINST         0\n",
       "HHDFMX         0\n",
       "HHDREL         0\n",
       "MIGSAME        0\n",
       "NOEMP          0\n",
       "PARENT         0\n",
       "PEFNTVTY       0\n",
       "PEMNTVTY       0\n",
       "PENATVTY       0\n",
       "PRCITSHP       0\n",
       "SEOTR          0\n",
       "VETQVA         0\n",
       "VETYN          0\n",
       "WKSWORK        0\n",
       "YEAR           0\n",
       "AAGE_bin       0\n",
       "ADTIND_bin     0\n",
       "ADTOCC_bin     0\n",
       "AHRSPAY_bin    0\n",
       "CAPGAIN_bin    0\n",
       "CAPLOSS_bin    0\n",
       "DIVVAL_bin     0\n",
       "WKSWORK_bin    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
       "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
       "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
       "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# selecting columns with object datatypes\n",
    "df_more_than_50k_obj = df_more_than_50k.select_dtypes(include=[np.object])\n",
    "df_more_than_50k_obj.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bucketization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The categorical values of each of the categorical attributes were studied and these appear to be different ,\n",
    "# hence a logical bucketization appeared not to be possible and therefore was not performed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# One-Hot Encoding: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      ACLSWKR_ Federal government  ACLSWKR_ Local government  \\\n",
      "0                               0                          0   \n",
      "1                               0                          0   \n",
      "2                               0                          0   \n",
      "3                               0                          0   \n",
      "4                               0                          0   \n",
      "...                           ...                        ...   \n",
      "3575                            0                          0   \n",
      "3576                            0                          0   \n",
      "3577                            0                          0   \n",
      "3578                            0                          0   \n",
      "3579                            0                          0   \n",
      "\n",
      "      ACLSWKR_ Never worked  ACLSWKR_ Not in universe  ACLSWKR_ Private  \\\n",
      "0                         0                         0                 0   \n",
      "1                         0                         0                 1   \n",
      "2                         0                         0                 1   \n",
      "3                         0                         0                 1   \n",
      "4                         0                         0                 1   \n",
      "...                     ...                       ...               ...   \n",
      "3575                      0                         0                 1   \n",
      "3576                      0                         0                 0   \n",
      "3577                      0                         0                 1   \n",
      "3578                      0                         0                 0   \n",
      "3579                      0                         1                 0   \n",
      "\n",
      "      ACLSWKR_ Self-employed-incorporated  \\\n",
      "0                                       1   \n",
      "1                                       0   \n",
      "2                                       0   \n",
      "3                                       0   \n",
      "4                                       0   \n",
      "...                                   ...   \n",
      "3575                                    0   \n",
      "3576                                    0   \n",
      "3577                                    0   \n",
      "3578                                    1   \n",
      "3579                                    0   \n",
      "\n",
      "      ACLSWKR_ Self-employed-not incorporated  ACLSWKR_ State government  \\\n",
      "0                                           0                          0   \n",
      "1                                           0                          0   \n",
      "2                                           0                          0   \n",
      "3                                           0                          0   \n",
      "4                                           0                          0   \n",
      "...                                       ...                        ...   \n",
      "3575                                        0                          0   \n",
      "3576                                        1                          0   \n",
      "3577                                        0                          0   \n",
      "3578                                        0                          0   \n",
      "3579                                        0                          0   \n",
      "\n",
      "      AHGA_ 10th grade  AHGA_ 11th grade  ...  PENATVTY_ Vietnam  \\\n",
      "0                    0                 0  ...                  0   \n",
      "1                    0                 0  ...                  0   \n",
      "2                    0                 0  ...                  0   \n",
      "3                    0                 0  ...                  0   \n",
      "4                    0                 0  ...                  0   \n",
      "...                ...               ...  ...                ...   \n",
      "3575                 0                 0  ...                  0   \n",
      "3576                 0                 0  ...                  0   \n",
      "3577                 0                 0  ...                  0   \n",
      "3578                 0                 0  ...                  0   \n",
      "3579                 0                 0  ...                  0   \n",
      "\n",
      "      PENATVTY_ Yugoslavia  PRCITSHP_ Foreign born- Not a citizen of U S   \\\n",
      "0                        0                                              0   \n",
      "1                        0                                              0   \n",
      "2                        0                                              0   \n",
      "3                        0                                              0   \n",
      "4                        0                                              0   \n",
      "...                    ...                                            ...   \n",
      "3575                     0                                              0   \n",
      "3576                     0                                              0   \n",
      "3577                     0                                              0   \n",
      "3578                     0                                              0   \n",
      "3579                     0                                              0   \n",
      "\n",
      "      PRCITSHP_ Foreign born- U S citizen by naturalization  \\\n",
      "0                                                     0       \n",
      "1                                                     0       \n",
      "2                                                     0       \n",
      "3                                                     0       \n",
      "4                                                     0       \n",
      "...                                                 ...       \n",
      "3575                                                  0       \n",
      "3576                                                  0       \n",
      "3577                                                  0       \n",
      "3578                                                  0       \n",
      "3579                                                  0       \n",
      "\n",
      "      PRCITSHP_ Native- Born abroad of American Parent(s)  \\\n",
      "0                                                     0     \n",
      "1                                                     0     \n",
      "2                                                     0     \n",
      "3                                                     0     \n",
      "4                                                     0     \n",
      "...                                                 ...     \n",
      "3575                                                  0     \n",
      "3576                                                  0     \n",
      "3577                                                  0     \n",
      "3578                                                  0     \n",
      "3579                                                  0     \n",
      "\n",
      "      PRCITSHP_ Native- Born in Puerto Rico or U S Outlying  \\\n",
      "0                                                     0       \n",
      "1                                                     0       \n",
      "2                                                     0       \n",
      "3                                                     0       \n",
      "4                                                     0       \n",
      "...                                                 ...       \n",
      "3575                                                  0       \n",
      "3576                                                  0       \n",
      "3577                                                  0       \n",
      "3578                                                  0       \n",
      "3579                                                  0       \n",
      "\n",
      "      PRCITSHP_ Native- Born in the United States  VETQVA_ No  \\\n",
      "0                                               1           0   \n",
      "1                                               1           1   \n",
      "2                                               1           0   \n",
      "3                                               1           0   \n",
      "4                                               1           0   \n",
      "...                                           ...         ...   \n",
      "3575                                            1           0   \n",
      "3576                                            1           0   \n",
      "3577                                            1           0   \n",
      "3578                                            1           0   \n",
      "3579                                            1           1   \n",
      "\n",
      "      VETQVA_ Not in universe  VETQVA_ Yes  \n",
      "0                           1            0  \n",
      "1                           0            0  \n",
      "2                           1            0  \n",
      "3                           1            0  \n",
      "4                           1            0  \n",
      "...                       ...          ...  \n",
      "3575                        1            0  \n",
      "3576                        1            0  \n",
      "3577                        1            0  \n",
      "3578                        1            0  \n",
      "3579                        0            0  \n",
      "\n",
      "[3580 rows x 302 columns]\n"
     ]
    }
   ],
   "source": [
    "# one hot encode the required features\n",
    "df_more_than_50k_one_hot_encoded_data = pd.get_dummies(df_more_than_50k[['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
    "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
    "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
    "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA']], columns = ['ACLSWKR', 'AHGA', 'AHSCOL', 'AMARITL', 'AMJIND', 'AMJOCC', 'ARACE',\n",
    "       'AREORGN', 'ASEX', 'AUNMEM', 'AUNTYPE', 'AWKSTAT', 'FILESTAT',\n",
    "       'GRINREG', 'GRINST', 'HHDFMX', 'HHDREL', 'MIGSAME', 'PARENT',\n",
    "       'PEFNTVTY', 'PEMNTVTY', 'PENATVTY', 'PRCITSHP', 'VETQVA'])\n",
    "print(df_more_than_50k_one_hot_encoded_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ACLSWKR_ Federal government', 'ACLSWKR_ Local government',\n",
       "       'ACLSWKR_ Never worked', 'ACLSWKR_ Not in universe', 'ACLSWKR_ Private',\n",
       "       'ACLSWKR_ Self-employed-incorporated',\n",
       "       'ACLSWKR_ Self-employed-not incorporated', 'ACLSWKR_ State government',\n",
       "       'AHGA_ 10th grade', 'AHGA_ 11th grade',\n",
       "       ...\n",
       "       'PENATVTY_ Vietnam', 'PENATVTY_ Yugoslavia',\n",
       "       'PRCITSHP_ Foreign born- Not a citizen of U S ',\n",
       "       'PRCITSHP_ Foreign born- U S citizen by naturalization',\n",
       "       'PRCITSHP_ Native- Born abroad of American Parent(s)',\n",
       "       'PRCITSHP_ Native- Born in Puerto Rico or U S Outlying',\n",
       "       'PRCITSHP_ Native- Born in the United States', 'VETQVA_ No',\n",
       "       'VETQVA_ Not in universe', 'VETQVA_ Yes'],\n",
       "      dtype='object', length=302)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_more_than_50k_one_hot_encoded_data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "# selecting columns with integer datatypes\n",
    "df_more_than_50k_int =df_more_than_50k[['AAGE_bin', 'ADTOCC_bin', 'ADTIND_bin',\n",
    "       'AHRSPAY_bin', 'CAPGAIN_bin', 'CAPLOSS_bin', 'DIVVAL_bin', 'NOEMP',\n",
    "       'SEOTR', 'VETYN', 'WKSWORK_bin', 'YEAR']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>AAGE_bin</th>\n",
       "      <th>ADTOCC_bin</th>\n",
       "      <th>ADTIND_bin</th>\n",
       "      <th>AHRSPAY_bin</th>\n",
       "      <th>CAPGAIN_bin</th>\n",
       "      <th>CAPLOSS_bin</th>\n",
       "      <th>DIVVAL_bin</th>\n",
       "      <th>NOEMP</th>\n",
       "      <th>SEOTR</th>\n",
       "      <th>VETYN</th>\n",
       "      <th>...</th>\n",
       "      <th>PENATVTY_ Vietnam</th>\n",
       "      <th>PENATVTY_ Yugoslavia</th>\n",
       "      <th>PRCITSHP_ Foreign born- Not a citizen of U S</th>\n",
       "      <th>PRCITSHP_ Foreign born- U S citizen by naturalization</th>\n",
       "      <th>PRCITSHP_ Native- Born abroad of American Parent(s)</th>\n",
       "      <th>PRCITSHP_ Native- Born in Puerto Rico or U S Outlying</th>\n",
       "      <th>PRCITSHP_ Native- Born in the United States</th>\n",
       "      <th>VETQVA_ No</th>\n",
       "      <th>VETQVA_ Not in universe</th>\n",
       "      <th>VETQVA_ Yes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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",
       "      <td>...</td>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3575</th>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3576</th>\n",
       "      <td>7</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3577</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3578</th>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3579</th>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3580 rows × 314 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     AAGE_bin ADTOCC_bin ADTIND_bin AHRSPAY_bin CAPGAIN_bin CAPLOSS_bin  \\\n",
       "0           6          2          3           1           1           1   \n",
       "1           4          7          2           1           1           1   \n",
       "2           6          5          3           1           1           1   \n",
       "3           1          2          5           1           2           1   \n",
       "4           1          2          3           1           1           1   \n",
       "...       ...        ...        ...         ...         ...         ...   \n",
       "3575        6          7          3           1           1           1   \n",
       "3576        7          3          6           1           1           1   \n",
       "3577        3          3          6           1           1           1   \n",
       "3578        6          2          5           1           1           2   \n",
       "3579        7          1          1           1           1           2   \n",
       "\n",
       "     DIVVAL_bin  NOEMP  SEOTR  VETYN  ... PENATVTY_ Vietnam  \\\n",
       "0             2      2      0      2  ...                 0   \n",
       "1             1      4      0      1  ...                 0   \n",
       "2             1      6      0      2  ...                 0   \n",
       "3             2      2      1      2  ...                 0   \n",
       "4             2      6      0      2  ...                 0   \n",
       "...         ...    ...    ...    ...  ...               ...   \n",
       "3575          2      6      0      2  ...                 0   \n",
       "3576          1      1      0      2  ...                 0   \n",
       "3577          1      6      0      2  ...                 0   \n",
       "3578          2      2      0      2  ...                 0   \n",
       "3579          1      0      0      1  ...                 0   \n",
       "\n",
       "      PENATVTY_ Yugoslavia  PRCITSHP_ Foreign born- Not a citizen of U S   \\\n",
       "0                        0                                              0   \n",
       "1                        0                                              0   \n",
       "2                        0                                              0   \n",
       "3                        0                                              0   \n",
       "4                        0                                              0   \n",
       "...                    ...                                            ...   \n",
       "3575                     0                                              0   \n",
       "3576                     0                                              0   \n",
       "3577                     0                                              0   \n",
       "3578                     0                                              0   \n",
       "3579                     0                                              0   \n",
       "\n",
       "      PRCITSHP_ Foreign born- U S citizen by naturalization  \\\n",
       "0                                                     0       \n",
       "1                                                     0       \n",
       "2                                                     0       \n",
       "3                                                     0       \n",
       "4                                                     0       \n",
       "...                                                 ...       \n",
       "3575                                                  0       \n",
       "3576                                                  0       \n",
       "3577                                                  0       \n",
       "3578                                                  0       \n",
       "3579                                                  0       \n",
       "\n",
       "      PRCITSHP_ Native- Born abroad of American Parent(s)  \\\n",
       "0                                                     0     \n",
       "1                                                     0     \n",
       "2                                                     0     \n",
       "3                                                     0     \n",
       "4                                                     0     \n",
       "...                                                 ...     \n",
       "3575                                                  0     \n",
       "3576                                                  0     \n",
       "3577                                                  0     \n",
       "3578                                                  0     \n",
       "3579                                                  0     \n",
       "\n",
       "      PRCITSHP_ Native- Born in Puerto Rico or U S Outlying  \\\n",
       "0                                                     0       \n",
       "1                                                     0       \n",
       "2                                                     0       \n",
       "3                                                     0       \n",
       "4                                                     0       \n",
       "...                                                 ...       \n",
       "3575                                                  0       \n",
       "3576                                                  0       \n",
       "3577                                                  0       \n",
       "3578                                                  0       \n",
       "3579                                                  0       \n",
       "\n",
       "      PRCITSHP_ Native- Born in the United States  VETQVA_ No  \\\n",
       "0                                               1           0   \n",
       "1                                               1           1   \n",
       "2                                               1           0   \n",
       "3                                               1           0   \n",
       "4                                               1           0   \n",
       "...                                           ...         ...   \n",
       "3575                                            1           0   \n",
       "3576                                            1           0   \n",
       "3577                                            1           0   \n",
       "3578                                            1           0   \n",
       "3579                                            1           1   \n",
       "\n",
       "      VETQVA_ Not in universe  VETQVA_ Yes  \n",
       "0                           1            0  \n",
       "1                           0            0  \n",
       "2                           1            0  \n",
       "3                           1            0  \n",
       "4                           1            0  \n",
       "...                       ...          ...  \n",
       "3575                        1            0  \n",
       "3576                        1            0  \n",
       "3577                        1            0  \n",
       "3578                        1            0  \n",
       "3579                        0            0  \n",
       "\n",
       "[3580 rows x 314 columns]"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create concatenated dataframe\n",
    "data_more_than_50k = pd.concat([df_more_than_50k_int, df_more_than_50k_one_hot_encoded_data], axis=1) \n",
    "data_more_than_50k"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_more_than_50k_1 = data_more_than_50k.copy()\n",
    "data_more_than_50k_1['Label'] = -1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performing Principal Component Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=2)\n",
    "data_more_than_50k_pca = pca.fit_transform(data_more_than_50k)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Finding the average within-cluster distance for the number of clusters (k) in the range 10 to 25"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Create empty lists to store cluster size and average within-cluster distances\n",
    "ks = []\n",
    "avg_dist = []\n",
    "\n",
    "# The kmedians function requires the initial value of medians to be given to it as an input.\n",
    "# However, no such value are given in the data, hence, we have used kmeans where k is ranging from 10 to 24 to find out\n",
    "# the set of centroids per value of k. These centroid values have been given as input to the kmedians function assuming\n",
    "# the centroids in the kmeans are the most central points in the clusters formed.\n",
    "\n",
    "for k in range(10,25):\n",
    "    model = KMeans(n_clusters=k, random_state=0).fit(data_more_than_50k_pca)\n",
    "    centers = np.array(model.cluster_centers_)\n",
    "    ks.append(k)\n",
    "    # Create instance of K-Medians algorithm.\n",
    "    kmedians_instance = kmedians(data_more_than_50k_pca, initial_medians = centers)\n",
    "    # Run cluster analysis and obtain results.\n",
    "    kmedians_instance.process()\n",
    "    clusters = kmedians_instance.get_clusters()\n",
    "    medians = kmedians_instance.get_medians()\n",
    "    for i in range(len(clusters)):\n",
    "        data_more_than_50k_1.iloc[data_more_than_50k_1.index.isin(clusters[i]),-1]=i\n",
    "    new_df_more_than_50k = pd.DataFrame(data_more_than_50k_pca, columns=['PC1','PC2'])\n",
    "    new_df_more_than_50k['Label'] = data_more_than_50k_1['Label']\n",
    "    dists=[]\n",
    "    for i, (cx, cy) in enumerate(medians):\n",
    "        a= pd.DataFrame()\n",
    "        a=new_df_more_than_50k[new_df_more_than_50k['Label'] == i].iloc[:, :-1]\n",
    "        a['dist'] = np.sqrt((a['PC1']-medians[i][0])**2+(a['PC2']-medians[i][1])**2)\n",
    "        dists.append(np.mean(a['dist']))\n",
    "    avg_dist.append(np.mean(dists))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plotting the average within-cluster distance vs. a number of clusters graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting the average within-cluster distance vs. a number of clusters graph.\n",
    "plt.plot(ks, avg_dist, 'bx-') \n",
    "plt.xlabel('Values of K') \n",
    "plt.ylabel('average within-cluster distance') \n",
    "plt.title('the average within-cluster distance vs. a number of clusters graph.') \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3 (f)\n",
    "\n",
    "# Compare the Population dataset with the more than 50k dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compare the proportion of data in each cluster for the more than 50k data to the proportion of data in each cluster for the general population."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = KMeans(n_clusters=22, random_state=0).fit(data_population_pca)\n",
    "centers = np.array(model.cluster_centers_)\n",
    "ks.append(k)\n",
    "# Create instance of K-Medians algorithm.\n",
    "kmedians_instance1 = kmedians(data_population_pca, initial_medians = centers)\n",
    "# Run cluster analysis and obtain results.\n",
    "kmedians_instance1.process()\n",
    "clusters = kmedians_instance1.get_clusters()\n",
    "medians = kmedians_instance1.get_medians()\n",
    "for i in range(len(clusters)):\n",
    "    data_population_1.iloc[data_population_1.index.isin(clusters[i]),-1]=i\n",
    "new_df_population = pd.DataFrame(data_population_pca, columns=['PC1','PC2'])\n",
    "new_df_population['Label'] = data_population_1['Label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = KMeans(n_clusters=22, random_state=0).fit(data_more_than_50k_pca)\n",
    "centers = np.array(model.cluster_centers_)\n",
    "ks.append(k)\n",
    "# Create instance of K-Medians algorithm.\n",
    "kmedians_instance2 = kmedians(data_more_than_50k_pca, initial_medians = centers)\n",
    "# Run cluster analysis and obtain results.\n",
    "kmedians_instance2.process()\n",
    "clusters = kmedians_instance2.get_clusters()\n",
    "medians = kmedians_instance2.get_medians()\n",
    "for i in range(len(clusters)):\n",
    "    data_more_than_50k_1.iloc[data_more_than_50k_1.index.isin(clusters[i]),-1]=i\n",
    "new_df_more_than_50k = pd.DataFrame(data_more_than_50k_pca, columns=['PC1','PC2'])\n",
    "new_df_more_than_50k['Label'] = data_more_than_50k_1['Label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "# proportion of data in each cluster for the more than 50k data and for the general population\n",
    "population_cluster_prop = new_df_population['Label'].value_counts()/new_df_population.shape[0]\n",
    "more_than_50k_cluster_prop = new_df_more_than_50k['Label'].value_counts()/new_df_more_than_50k.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# barplots to show proportion of data in each cluster in the general population and more than 50k data\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize=(30, 8))\n",
    "fig.suptitle('Cluster PROPORTION')\n",
    "ax2.set_title('more_than_50k_cluster_prop')\n",
    "ax1.set_title('General Population')\n",
    "sns.barplot(x=more_than_50k_cluster_prop.index, y=more_than_50k_cluster_prop, ax=ax2)\n",
    "sns.barplot(x=population_cluster_prop.index, y=population_cluster_prop, ax=ax1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## What kind of people are part of a cluster overrepresented in the more than 50k data compared to the general population?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Our strategy to find out overreprentation is to sort the clusters in decreasing order of their counts and\n",
    "# compare the largest cluster of population.csv to that of more_than_50k"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Finding topmost cluster represented"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9     0.081685\n",
       "14    0.079570\n",
       "3     0.074307\n",
       "12    0.071470\n",
       "2     0.064754\n",
       "19    0.055332\n",
       "11    0.054510\n",
       "0     0.044636\n",
       "7     0.044361\n",
       "8     0.043820\n",
       "21    0.039955\n",
       "10    0.039690\n",
       "1     0.037484\n",
       "20    0.037324\n",
       "4     0.035650\n",
       "18    0.035409\n",
       "6     0.033816\n",
       "5     0.030829\n",
       "13    0.029906\n",
       "16    0.028984\n",
       "15    0.019682\n",
       "17    0.016825\n",
       "Name: Label, dtype: float64"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# population dataset\n",
    "cluster_population = population_cluster_prop.sort_values(ascending=False)\n",
    "cluster_population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3     0.070112\n",
       "4     0.070112\n",
       "11    0.066480\n",
       "21    0.054749\n",
       "10    0.051676\n",
       "1     0.050279\n",
       "5     0.050000\n",
       "19    0.049162\n",
       "16    0.048045\n",
       "2     0.048045\n",
       "14    0.046927\n",
       "18    0.044413\n",
       "6     0.044134\n",
       "13    0.042179\n",
       "17    0.040782\n",
       "15    0.040223\n",
       "12    0.037151\n",
       "7     0.036313\n",
       "20    0.035475\n",
       "8     0.030447\n",
       "9     0.026536\n",
       "0     0.016760\n",
       "Name: Label, dtype: float64"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# more_than_50k dataset\n",
    "cluster_more_than_50k = more_than_50k_cluster_prop.sort_values(ascending=False)\n",
    "cluster_more_than_50k"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Applying PCA inverse transform to the features of the topmost cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The strategy is as follows: We assume that the overrepresented cluster is the one that is having the maximum count for each of the \n",
    "# dataset respectively\n",
    "# As the cluster number are automatically assigned by the function . hence direct differences between similar numbered\n",
    "# clusters of the two datasets seems to be illogical. Hence we have tried to avoid this."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# We perform inverse transform of PCA to find out the most impactful features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_scaled_comparison(df_sample, kmedians_instance, cluster, pca):\n",
    "    #pca inverse transform on clusters\n",
    "    m = kmedians_instance.get_medians()\n",
    "    inv_pca= pca.inverse_transform(m[cluster])\n",
    "    #convert to dict and use columns from full space\n",
    "    df_dict=dict(zip(df_sample.columns, inv_pca))\n",
    "    #make dataframe\n",
    "    df=pd.DataFrame.from_dict(df_dict, orient='index').rename( \n",
    "        columns={0: 'feature_values'}).sort_values('feature_values', ascending=False)\n",
    "    df['feature_values_abs']=abs(df['feature_values'])\n",
    "    viz_df= pd.concat((df['feature_values'][:10], df['feature_values'][-10:]), axis=0)\n",
    "    viz_df.plot(kind='barh')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_scaled_comparison(data_population, kmedians_instance1,9, pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Description for top 5 features in the population data \n",
    "# YEAR: data collected in the year\n",
    "# NOEMP : num persons worked for employer\n",
    "# ADTIND_bin: industry code\n",
    "# WKSWORK_bin: weeks worked in year\n",
    "# ADTOCC_bin: occupation code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filtering records for population data\n",
    "ind1 = new_df_population[new_df_population['Label'] ==9].index\n",
    "data_population_clus1 = data_population[new_df_population.index.isin(ind1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min data collected in the year : 94\n",
      " Max data collected in the year : 95\n",
      " Mean data collected in the year : 94.49306663394282\n",
      " Median data collected in the year : 94.0\n",
      " Mode data collected in the year : 0    94\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\" Min data collected in the year :\", data_population_clus1['YEAR'].min())\n",
    "print(\" Max data collected in the year :\", data_population_clus1['YEAR'].max())\n",
    "print(\" Mean data collected in the year :\", data_population_clus1['YEAR'].mean())\n",
    "print(\" Median data collected in the year :\", data_population_clus1['YEAR'].median())\n",
    "print(\" Mode data collected in the year :\", data_population_clus1['YEAR'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min num persons worked for employer : 0\n",
      " Max num persons worked for employer : 0\n",
      " Mean num persons worked for employer : 0.0\n",
      " Median num persons worked for employer : 0.0\n",
      " Mode num persons worked for employer : 0    0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\" Min num persons worked for employer :\", data_population_clus1['NOEMP'].min())\n",
    "print(\" Max num persons worked for employer :\", data_population_clus1['NOEMP'].max())\n",
    "print(\" Mean num persons worked for employer :\", data_population_clus1['NOEMP'].mean())\n",
    "print(\" Median num persons worked for employer :\", data_population_clus1['NOEMP'].median())\n",
    "print(\" Mode num persons worked for employer :\", data_population_clus1['NOEMP'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Minimum and maximum number persons worked for employer are 0 and 0\n",
    "# Mean of number persons worked for employer is 0\n",
    "# mode of the number persons worked for employer are 0 \n",
    "# median of the number persons worked for employer are 0 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min industry code : 1\n",
      " Max industry code : 1\n",
      " Mode industry code : 0    1\n",
      "Name: ADTIND_bin, dtype: category\n",
      "Categories (7, object): [1 < 2 < 3 < 4 < 5 < 6 < 7]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min industry code :\", data_population_clus1['ADTIND_bin'].min())\n",
    "print(\" Max industry code :\", data_population_clus1['ADTIND_bin'].max())\n",
    "print(\" Mode industry code :\", data_population_clus1['ADTIND_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum, maximum  and mode ADTIND_bin refers to industry code of 1,1  and 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min weeks worked in year  : 1\n",
      " Max weeks worked in year  : 1\n",
      " Mode weeks worked in year : 0    1\n",
      "Name: WKSWORK_bin, dtype: category\n",
      "Categories (4, object): [1 < 2 < 3 < 4]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min weeks worked in year  :\", data_population_clus1['WKSWORK_bin'].min())\n",
    "print(\" Max weeks worked in year  :\", data_population_clus1['WKSWORK_bin'].max())\n",
    "print(\" Mode weeks worked in year :\", data_population_clus1['WKSWORK_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum WKSWORK_bin of 1 and 1 refers to 1 weeks worked in year "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min occupation code : 1\n",
      " Max occupation code : 1\n",
      " Mode occupation code: 0    1\n",
      "Name: ADTOCC_bin, dtype: category\n",
      "Categories (7, object): [1 < 2 < 3 < 4 < 5 < 6 < 7]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min occupation code :\", data_population_clus1['ADTOCC_bin'].min())\n",
    "print(\" Max occupation code :\", data_population_clus1['ADTOCC_bin'].max())\n",
    "print(\" Mode occupation code:\", data_population_clus1['ADTOCC_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum ADTOCC_bin of 1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_scaled_comparison(data_more_than_50k, kmedians_instance2, 3, pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Description for top 5 features in the more_than_50k data\n",
    "# YEAR: data collected in the year\n",
    "# NOEMP: num persons worked for employer\n",
    "# ADTIND_bin: industry code\n",
    "# WKSWORK_bin: weeks worked in year\n",
    "# AAGE_bin: age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filtering records for more_than_50k data\n",
    "ind2 = new_df_more_than_50k[new_df_more_than_50k['Label'] ==3].index\n",
    "data_more_than_50k_clus2 = data_more_than_50k[new_df_more_than_50k.index.isin(ind2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min data collected in the year : 94\n",
      " Max data collected in the year : 95\n",
      " Mean data collected in the year : 94.5219123505976\n",
      " Median data collected in the year : 95.0\n",
      " Mode data collected in the year : 0    95\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\" Min data collected in the year :\", data_more_than_50k_clus2['YEAR'].min())\n",
    "print(\" Max data collected in the year :\", data_more_than_50k_clus2['YEAR'].max())\n",
    "print(\" Mean data collected in the year :\", data_more_than_50k_clus2['YEAR'].mean())\n",
    "print(\" Median data collected in the year :\", data_more_than_50k_clus2['YEAR'].median())\n",
    "print(\" Mode data collected in the year :\", data_more_than_50k_clus2['YEAR'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum data collected in the year are 94 and 95 with mean of 94.5219123505976 and median of 94.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min num persons worked for employer : 3\n",
      " Max num persons worked for employer : 6\n",
      " Mean num persons worked for employer : 5.334661354581673\n",
      " Median num persons worked for employer : 6.0\n",
      " Mode num persons worked for employer : 0    6\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\" Min num persons worked for employer :\", data_more_than_50k_clus2['NOEMP'].min())\n",
    "print(\" Max num persons worked for employer :\", data_more_than_50k_clus2['NOEMP'].max())\n",
    "print(\" Mean num persons worked for employer :\", data_more_than_50k_clus2['NOEMP'].mean())\n",
    "print(\" Median num persons worked for employer :\", data_more_than_50k_clus2['NOEMP'].median())\n",
    "print(\" Mode num persons worked for employer :\", data_more_than_50k_clus2['NOEMP'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Minimum and maximum number persons worked for employer are 3 and 6\n",
    "# Mean of number persons worked for employer is 5.334661354581673\n",
    "# mode of the number persons worked for employer are 6 \n",
    "# median of the number persons worked for employer are 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min industry code : 4\n",
      " Max industry code : 6\n",
      " Mode industry code : 0    5\n",
      "Name: ADTIND_bin, dtype: category\n",
      "Categories (7, object): [1 < 2 < 3 < 4 < 5 < 6 < 7]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min industry code :\", data_more_than_50k_clus2['ADTIND_bin'].min())\n",
    "print(\" Max industry code :\", data_more_than_50k_clus2['ADTIND_bin'].max())\n",
    "print(\" Mode industry code :\", data_more_than_50k_clus2['ADTIND_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum ADTIND_bin refers to industry code of 4 and 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min weeks worked in year  : 2\n",
      " Max weeks worked in year  : 4\n",
      " Mode weeks worked in year : 0    4\n",
      "Name: WKSWORK_bin, dtype: category\n",
      "Categories (4, object): [1 < 2 < 3 < 4]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min weeks worked in year  :\", data_more_than_50k_clus2['WKSWORK_bin'].min())\n",
    "print(\" Max weeks worked in year  :\", data_more_than_50k_clus2['WKSWORK_bin'].max())\n",
    "print(\" Mode weeks worked in year :\", data_more_than_50k_clus2['WKSWORK_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum WKSWORK_bin of 2 and 4 refers to 1 to 52 weeks worked in year "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min age  : 1\n",
      " Max age  : 7\n",
      " Mode age : 0    5\n",
      "Name: AAGE_bin, dtype: category\n",
      "Categories (8, object): [1 < 2 < 3 < 4 < 5 < 6 < 7 < 8]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min age  :\", data_more_than_50k_clus2['AAGE_bin'].min())\n",
    "print(\" Max age  :\", data_more_than_50k_clus2['AAGE_bin'].max())\n",
    "print(\" Mode age :\", data_more_than_50k_clus2['AAGE_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum AAGE_bin of 1 and 7 refers to age between 18 to 70"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Clusters that are over-represented in the general population vs more than 50k population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "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>proportion_population</th>\n",
       "      <th>cluster_num_population</th>\n",
       "      <th>proportion_more_than_50k</th>\n",
       "      <th>cluster_num_more_than_50k</th>\n",
       "      <th>Difference</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.081685</td>\n",
       "      <td>9</td>\n",
       "      <td>0.070112</td>\n",
       "      <td>3</td>\n",
       "      <td>0.011573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.079570</td>\n",
       "      <td>14</td>\n",
       "      <td>0.070112</td>\n",
       "      <td>4</td>\n",
       "      <td>0.009458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.074307</td>\n",
       "      <td>3</td>\n",
       "      <td>0.066480</td>\n",
       "      <td>11</td>\n",
       "      <td>0.007827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.071470</td>\n",
       "      <td>12</td>\n",
       "      <td>0.054749</td>\n",
       "      <td>21</td>\n",
       "      <td>0.016722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.064754</td>\n",
       "      <td>2</td>\n",
       "      <td>0.051676</td>\n",
       "      <td>10</td>\n",
       "      <td>0.013078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.055332</td>\n",
       "      <td>19</td>\n",
       "      <td>0.050279</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.054510</td>\n",
       "      <td>11</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>5</td>\n",
       "      <td>0.004510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.044636</td>\n",
       "      <td>0</td>\n",
       "      <td>0.049162</td>\n",
       "      <td>19</td>\n",
       "      <td>-0.004526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.044361</td>\n",
       "      <td>7</td>\n",
       "      <td>0.048045</td>\n",
       "      <td>16</td>\n",
       "      <td>-0.003684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.043820</td>\n",
       "      <td>8</td>\n",
       "      <td>0.048045</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.004225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.039955</td>\n",
       "      <td>21</td>\n",
       "      <td>0.046927</td>\n",
       "      <td>14</td>\n",
       "      <td>-0.006972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.039690</td>\n",
       "      <td>10</td>\n",
       "      <td>0.044413</td>\n",
       "      <td>18</td>\n",
       "      <td>-0.004724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.037484</td>\n",
       "      <td>1</td>\n",
       "      <td>0.044134</td>\n",
       "      <td>6</td>\n",
       "      <td>-0.006650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.037324</td>\n",
       "      <td>20</td>\n",
       "      <td>0.042179</td>\n",
       "      <td>13</td>\n",
       "      <td>-0.004855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.035650</td>\n",
       "      <td>4</td>\n",
       "      <td>0.040782</td>\n",
       "      <td>17</td>\n",
       "      <td>-0.005132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.035409</td>\n",
       "      <td>18</td>\n",
       "      <td>0.040223</td>\n",
       "      <td>15</td>\n",
       "      <td>-0.004814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.033816</td>\n",
       "      <td>6</td>\n",
       "      <td>0.037151</td>\n",
       "      <td>12</td>\n",
       "      <td>-0.003335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.030829</td>\n",
       "      <td>5</td>\n",
       "      <td>0.036313</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.005484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.029906</td>\n",
       "      <td>13</td>\n",
       "      <td>0.035475</td>\n",
       "      <td>20</td>\n",
       "      <td>-0.005569</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.028984</td>\n",
       "      <td>16</td>\n",
       "      <td>0.030447</td>\n",
       "      <td>8</td>\n",
       "      <td>-0.001463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.019682</td>\n",
       "      <td>15</td>\n",
       "      <td>0.026536</td>\n",
       "      <td>9</td>\n",
       "      <td>-0.006854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.016825</td>\n",
       "      <td>17</td>\n",
       "      <td>0.016760</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000065</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    proportion_population  cluster_num_population  proportion_more_than_50k  \\\n",
       "0                0.081685                       9                  0.070112   \n",
       "1                0.079570                      14                  0.070112   \n",
       "2                0.074307                       3                  0.066480   \n",
       "3                0.071470                      12                  0.054749   \n",
       "4                0.064754                       2                  0.051676   \n",
       "5                0.055332                      19                  0.050279   \n",
       "6                0.054510                      11                  0.050000   \n",
       "7                0.044636                       0                  0.049162   \n",
       "8                0.044361                       7                  0.048045   \n",
       "9                0.043820                       8                  0.048045   \n",
       "10               0.039955                      21                  0.046927   \n",
       "11               0.039690                      10                  0.044413   \n",
       "12               0.037484                       1                  0.044134   \n",
       "13               0.037324                      20                  0.042179   \n",
       "14               0.035650                       4                  0.040782   \n",
       "15               0.035409                      18                  0.040223   \n",
       "16               0.033816                       6                  0.037151   \n",
       "17               0.030829                       5                  0.036313   \n",
       "18               0.029906                      13                  0.035475   \n",
       "19               0.028984                      16                  0.030447   \n",
       "20               0.019682                      15                  0.026536   \n",
       "21               0.016825                      17                  0.016760   \n",
       "\n",
       "    cluster_num_more_than_50k  Difference  \n",
       "0                           3    0.011573  \n",
       "1                           4    0.009458  \n",
       "2                          11    0.007827  \n",
       "3                          21    0.016722  \n",
       "4                          10    0.013078  \n",
       "5                           1    0.005053  \n",
       "6                           5    0.004510  \n",
       "7                          19   -0.004526  \n",
       "8                          16   -0.003684  \n",
       "9                           2   -0.004225  \n",
       "10                         14   -0.006972  \n",
       "11                         18   -0.004724  \n",
       "12                          6   -0.006650  \n",
       "13                         13   -0.004855  \n",
       "14                         17   -0.005132  \n",
       "15                         15   -0.004814  \n",
       "16                         12   -0.003335  \n",
       "17                          7   -0.005484  \n",
       "18                         20   -0.005569  \n",
       "19                          8   -0.001463  \n",
       "20                          9   -0.006854  \n",
       "21                          0    0.000065  "
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating datadframe with proportion_ for population and num_more_than_50k clusters\n",
    "d1 = pd.DataFrame(cluster_population)\n",
    "d1['cluster_num'] = d1.index\n",
    "d1.columns=['proportion_population','cluster_num_population']\n",
    "d1.reset_index(drop=True, inplace=True)\n",
    "d2 = pd.DataFrame(cluster_more_than_50k)\n",
    "d2['cluster_num'] = d2.index\n",
    "d2.columns=['proportion_more_than_50k','cluster_num_more_than_50k']\n",
    "d2.reset_index(drop=True, inplace=True)\n",
    "d_concat = pd.concat([d1, d2], axis=1)\n",
    "# Difference between cluster proportion\n",
    "d_concat['Difference'] = d_concat['proportion_population'] - d_concat['proportion_more_than_50k']\n",
    "d_concat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Population dataset Over - represented compared to more_than_50k by 0.01157%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.00946%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.00783%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.01672%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.01308%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.00505%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.00451%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00453%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00368%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00423%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00697%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00472%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00665%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00485%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00513%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00481%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00334%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00548%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00557%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00146%\n",
      "Population dataset Under - represented compared to more_than_50k by -0.00685%\n",
      "Population dataset Over - represented compared to more_than_50k by 0.00007%\n"
     ]
    }
   ],
   "source": [
    "# To determine the overrepresentation the clusters formed by each of the two datasets have been sorted on the basis of \n",
    "# their percentage counts and the largest clusters have been chosen to find out the maximum overrepresentation. \n",
    "\n",
    "\n",
    "# Printing clusters that are over-represented in the general population vs more than 50k population\n",
    "\n",
    "for x in range(22):\n",
    "    if(d_concat['Difference'][x]>0):\n",
    "        print('Population dataset Over - represented compared to more_than_50k by %.5f%%'%(d_concat['Difference'][x]))\n",
    "    elif(d_concat['Difference'][x]<0):\n",
    "        print('Population dataset Under - represented compared to more_than_50k by %.5f%%'% ( d_concat['Difference'][x]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Analyze a cluster overrepresented in the more than 50k data compared to the general population."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cluster 9 is overrepresented in the more than 50k data compared to the general population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_scaled_comparison(df_sample, kmedians_instance, cluster, pca):\n",
    "    #pca inverse transform on clusters\n",
    "    m = kmedians_instance.get_medians()\n",
    "    inv_pca= pca.inverse_transform(m[cluster])\n",
    "    #convert to dict and use columns from full space\n",
    "    df_dict=dict(zip(df_sample.columns, inv_pca))\n",
    "    #make dataframe\n",
    "    df=pd.DataFrame.from_dict(df_dict, orient='index').rename( \n",
    "        columns={0: 'feature_values'}).sort_values('feature_values', ascending=False)\n",
    "    df['feature_values_abs']=abs(df['feature_values'])\n",
    "    viz_df= pd.concat((df['feature_values'][:10], df['feature_values'][-10:]), axis=0)\n",
    "    viz_df.plot(kind='barh')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_scaled_comparison(data_more_than_50k, kmedians_instance2, 9, pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Description for top 5 features in the more_than_50k data for cluster 9\n",
    "# YEAR: data collected in the year\n",
    "# ADTIND_bin: industry code \n",
    "# AAGE_bin: age\n",
    "# WKSWORK_bin: weeks worked in year\n",
    "# ADTOCC_bin: occupation code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filtering records for cluster 5\n",
    "ind = new_df_more_than_50k[new_df_more_than_50k['Label'] ==5].index\n",
    "data_more_than_50k_clus5 = data_more_than_50k[new_df_more_than_50k.index.isin(ind)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min data collected in the year : 94\n",
      " Max data collected in the year : 95\n",
      " Mean data collected in the year : 94.47486033519553\n",
      " Median data collected in the year : 94.0\n",
      " Mode data collected in the year : 0    94\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\" Min data collected in the year :\", data_more_than_50k_clus5['YEAR'].min())\n",
    "print(\" Max data collected in the year :\", data_more_than_50k_clus5['YEAR'].max())\n",
    "print(\" Mean data collected in the year :\", data_more_than_50k_clus5['YEAR'].mean())\n",
    "print(\" Median data collected in the year :\", data_more_than_50k_clus5['YEAR'].median())\n",
    "print(\" Mode data collected in the year :\", data_more_than_50k_clus5['YEAR'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min industry code : 1\n",
      " Max industry code : 1\n",
      " Mode industry code : 0    1\n",
      "Name: ADTIND_bin, dtype: category\n",
      "Categories (7, object): [1 < 2 < 3 < 4 < 5 < 6 < 7]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min industry code :\", data_population_clus1['ADTIND_bin'].min())\n",
    "print(\" Max industry code :\", data_population_clus1['ADTIND_bin'].max())\n",
    "print(\" Mode industry code :\", data_population_clus1['ADTIND_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum industry code are 1 and 1 with mode of1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min age  : 4\n",
      " Max age  : 8\n",
      " Mode age : 0    8\n",
      "Name: AAGE_bin, dtype: category\n",
      "Categories (8, object): [1 < 2 < 3 < 4 < 5 < 6 < 7 < 8]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min age  :\", data_more_than_50k_clus5['AAGE_bin'].min())\n",
    "print(\" Max age  :\", data_more_than_50k_clus5['AAGE_bin'].max())\n",
    "print(\" Mode age :\", data_more_than_50k_clus5['AAGE_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum AAGE_bin of 4 and 8 refers to age between 40 to 90"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min weeks worked in year  : 1\n",
      " Max weeks worked in year  : 4\n",
      " Mode weeks worked in year : 0    1\n",
      "Name: WKSWORK_bin, dtype: category\n",
      "Categories (4, object): [1 < 2 < 3 < 4]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min weeks worked in year  :\", data_more_than_50k_clus5['WKSWORK_bin'].min())\n",
    "print(\" Max weeks worked in year  :\", data_more_than_50k_clus5['WKSWORK_bin'].max())\n",
    "print(\" Mode weeks worked in year :\", data_more_than_50k_clus5['WKSWORK_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum WKSWORK_bin of 1 and 4 refers to 0 to 52 weeks worked in year "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Min occupation code : 1\n",
      " Max occupation code : 2\n",
      " Mode occupation code: 0    1\n",
      "Name: ADTOCC_bin, dtype: category\n",
      "Categories (7, object): [1 < 2 < 3 < 4 < 5 < 6 < 7]\n"
     ]
    }
   ],
   "source": [
    "print(\" Min occupation code :\", data_more_than_50k_clus5['ADTOCC_bin'].min())\n",
    "print(\" Max occupation code :\", data_more_than_50k_clus5['ADTOCC_bin'].max())\n",
    "print(\" Mode occupation code:\", data_more_than_50k_clus5['ADTOCC_bin'].mode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimum and maximum ADTOCC_bin of 1 and 4 refers to 1 and 2 occupation code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}