module 3/Data representation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# <img style=\"float: left; padding-right: 100px; width: 300px\" src=\"image/logo.png\">AI4SG Bootcamp:\n",
    "\n",
    "\n",
    "## Data representation and feature engineering \n",
    "**Authors:** Davis\n",
    "\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "np.random.seed(42)#for reproducible experiments\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "%matplotlib inline\n",
    "\n",
    "import warnings ## importing warnings library. \n",
    "warnings.filterwarnings('ignore') ## Ignore warning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_PATH = \"../data/students_exams_results.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(DATA_PATH) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>continue_drop</th>\n",
       "      <th>student_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>caste</th>\n",
       "      <th>mathematics_marks</th>\n",
       "      <th>english_marks</th>\n",
       "      <th>science_marks</th>\n",
       "      <th>science_teacher</th>\n",
       "      <th>languages_teacher</th>\n",
       "      <th>guardian</th>\n",
       "      <th>internet</th>\n",
       "      <th>school_id</th>\n",
       "      <th>total_students</th>\n",
       "      <th>total_toilets</th>\n",
       "      <th>establishment_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>continue</td>\n",
       "      <td>s17477</td>\n",
       "      <td>M</td>\n",
       "      <td>BC</td>\n",
       "      <td>0.393</td>\n",
       "      <td>0.338</td>\n",
       "      <td>0.393</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>mother</td>\n",
       "      <td>1</td>\n",
       "      <td>362</td>\n",
       "      <td>397</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1950.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>continue</td>\n",
       "      <td>s16612</td>\n",
       "      <td>M</td>\n",
       "      <td>SC</td>\n",
       "      <td>0.745</td>\n",
       "      <td>0.645</td>\n",
       "      <td>0.745</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>father</td>\n",
       "      <td>1</td>\n",
       "      <td>357</td>\n",
       "      <td>57</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1929.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>continue</td>\n",
       "      <td>s04010</td>\n",
       "      <td>M</td>\n",
       "      <td>BC</td>\n",
       "      <td>0.788</td>\n",
       "      <td>0.655</td>\n",
       "      <td>0.788</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>father</td>\n",
       "      <td>1</td>\n",
       "      <td>340</td>\n",
       "      <td>134</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1976.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>drop</td>\n",
       "      <td>s11124</td>\n",
       "      <td>F</td>\n",
       "      <td>BC</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.699</td>\n",
       "      <td>0.623</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>father</td>\n",
       "      <td>1</td>\n",
       "      <td>345</td>\n",
       "      <td>143</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1879.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>continue</td>\n",
       "      <td>s04384</td>\n",
       "      <td>M</td>\n",
       "      <td>SC</td>\n",
       "      <td>0.951</td>\n",
       "      <td>0.704</td>\n",
       "      <td>0.951</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>mother</td>\n",
       "      <td>1</td>\n",
       "      <td>304</td>\n",
       "      <td>390</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1914.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  continue_drop student_id gender caste  mathematics_marks  english_marks  \\\n",
       "0      continue     s17477      M    BC              0.393          0.338   \n",
       "1      continue     s16612      M    SC              0.745          0.645   \n",
       "2      continue     s04010      M    BC              0.788          0.655   \n",
       "3          drop     s11124      F    BC              0.623          0.699   \n",
       "4      continue     s04384      M    SC              0.951          0.704   \n",
       "\n",
       "   science_marks  science_teacher  languages_teacher guardian  internet  \\\n",
       "0          0.393                2                  2   mother         1   \n",
       "1          0.745                4                  3   father         1   \n",
       "2          0.788                8                  9   father         1   \n",
       "3          0.623                6                  0   father         1   \n",
       "4          0.951                8                  4   mother         1   \n",
       "\n",
       "   school_id  total_students  total_toilets  establishment_year  \n",
       "0        362             397            5.0              1950.0  \n",
       "1        357              57           14.0              1929.0  \n",
       "2        340             134           15.0              1976.0  \n",
       "3        345             143           28.0              1879.0  \n",
       "4        304             390           28.0              1914.0  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Total</th>\n",
       "      <th>Percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>establishment_year</th>\n",
       "      <td>309</td>\n",
       "      <td>1.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_toilets</th>\n",
       "      <td>309</td>\n",
       "      <td>1.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_students</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>school_id</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>internet</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>guardian</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>languages_teacher</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>science_teacher</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>science_marks</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>english_marks</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mathematics_marks</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>caste</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gender</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>student_id</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>continue_drop</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Total  Percent\n",
       "establishment_year    309      1.8\n",
       "total_toilets         309      1.8\n",
       "total_students          0      0.0\n",
       "school_id               0      0.0\n",
       "internet                0      0.0\n",
       "guardian                0      0.0\n",
       "languages_teacher       0      0.0\n",
       "science_teacher         0      0.0\n",
       "science_marks           0      0.0\n",
       "english_marks           0      0.0\n",
       "mathematics_marks       0      0.0\n",
       "caste                   0      0.0\n",
       "gender                  0      0.0\n",
       "student_id              0      0.0\n",
       "continue_drop           0      0.0"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total = data.isnull().sum().sort_values(ascending = False)\n",
    "percent = round(data.isnull().sum().sort_values(ascending = False)/len(data)*100, 2)\n",
    "missing_column=pd.concat([total, percent], axis = 1,keys= ['Total', 'Percent']) \n",
    "missing_column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One percent of data is missing in  total toilets and establishment year columns. Let ignore these columns by addding the following line in the load data function. We will also delete the student id column\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Question: What techniques can we use to handle this missing values?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will use this techniques later to address missing values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check data type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": [
    "data.dtypes.value_counts().sort_values().plot(kind='barh')\n",
    "plt.title('Number of columns distributed by Data Types')\n",
    "plt.xlabel('Number of columns')\n",
    "plt.ylabel('Data type');\n",
    "plt.savefig(\"data_distribution.pdf\", dpi=600)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us see how many categorical data do the columns having 'object' data types contain:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>student_id</th>\n",
       "      <td>17190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>guardian</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>caste</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gender</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>continue_drop</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Total\n",
       "student_id     17190\n",
       "guardian           4\n",
       "caste              4\n",
       "gender             2\n",
       "continue_drop      2"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "object_column=data.select_dtypes(include=['object']).columns\n",
    "category_num=data[object_column].apply(pd.Series.nunique, axis = 0).sort_values(ascending = False)\n",
    "category_num=pd.concat([category_num], axis = 1,keys= ['Total']) \n",
    "category_num"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will have to encode the columns having only 2 categorical data and one-hot encode columns with more than 2 categorical data. However the column with several categorical data need special attention. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The student_id column have more than 17190 with very less unique data for each category. \n",
    "We will drop this  column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(['student_id'], axis=1, inplace=True)   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### int data type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "int_column=data.select_dtypes(include=['int64']).columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>Total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>school_id</th>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>total_students</th>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>languages_teacher</th>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>science_teacher</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>internet</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Total\n",
       "school_id            100\n",
       "total_students        90\n",
       "languages_teacher     13\n",
       "science_teacher       10\n",
       "internet               2"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "int_num=data[int_column].apply(pd.Series.nunique, axis = 0).sort_values(ascending = False)\n",
    "int_num=pd.concat([int_num], axis = 1,keys= ['Total']) \n",
    "int_num"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>science_teacher</th>\n",
       "      <th>languages_teacher</th>\n",
       "      <th>internet</th>\n",
       "      <th>school_id</th>\n",
       "      <th>total_students</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>362</td>\n",
       "      <td>397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>357</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>340</td>\n",
       "      <td>134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>345</td>\n",
       "      <td>143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>304</td>\n",
       "      <td>390</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   science_teacher  languages_teacher  internet  school_id  total_students\n",
       "0                2                  2         1        362             397\n",
       "1                4                  3         1        357              57\n",
       "2                8                  9         1        340             134\n",
       "3                6                  0         1        345             143\n",
       "4                8                  4         1        304             390"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[int_column].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To decide wether these column are numeric or categorical let visualize their histogram"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[int_column].hist(grid=False);\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It clear with exception to internet column the rest of the column are numeric. We will therefore cahnge their data type to float."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_int_float(df):\n",
    "    return df.astype('float64')\n",
    "\n",
    "data['languages_teacher'] = convert_int_float(data['languages_teacher'])\n",
    "data['school_id'] = convert_int_float(data['school_id'])\n",
    "data['total_students'] = convert_int_float(data['total_students'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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": [
    "data.dtypes.value_counts().sort_values().plot(kind='barh')\n",
    "plt.title('Number of columns distributed by Data Types')\n",
    "plt.xlabel('Number of columns')\n",
    "plt.ylabel('Data type');\n",
    "plt.savefig(\"data_distribution.pdf\", dpi=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "numeric_column=data.select_dtypes(include=['float64']).columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[numeric_column].hist(grid=False);\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Process categorical data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['continue_drop', 'gender', 'caste', 'guardian', 'science_teacher', 'internet']"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "object_column=list(data.select_dtypes(include=['object']).columns)\n",
    "int_column=list(data.select_dtypes(include=['int64']).columns)\n",
    "\n",
    "categ_column = object_column + int_column\n",
    "categ_column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is no categorical column with missing value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total = data[categ_column].isnull().sum().sort_values(ascending = False)\n",
    "missing_column=pd.concat([total], axis = 1,keys= ['Total']) \n",
    "categ_column_with_missing = list(missing_column[missing_column['Total']>0].index)\n",
    "categ_column_with_missing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "categ_data = data[categ_column]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "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>continue_drop</th>\n",
       "      <th>gender</th>\n",
       "      <th>caste</th>\n",
       "      <th>guardian</th>\n",
       "      <th>science_teacher</th>\n",
       "      <th>internet</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>continue</td>\n",
       "      <td>M</td>\n",
       "      <td>BC</td>\n",
       "      <td>mother</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>continue</td>\n",
       "      <td>M</td>\n",
       "      <td>SC</td>\n",
       "      <td>father</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>continue</td>\n",
       "      <td>M</td>\n",
       "      <td>BC</td>\n",
       "      <td>father</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>drop</td>\n",
       "      <td>F</td>\n",
       "      <td>BC</td>\n",
       "      <td>father</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>continue</td>\n",
       "      <td>M</td>\n",
       "      <td>SC</td>\n",
       "      <td>mother</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  continue_drop gender caste guardian  science_teacher  internet\n",
       "0      continue      M    BC   mother                2         1\n",
       "1      continue      M    SC   father                4         1\n",
       "2      continue      M    BC   father                8         1\n",
       "3          drop      F    BC   father                6         1\n",
       "4      continue      M    SC   mother                8         1"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categ_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Skealrn offer several method for handling categorical data\n",
    "\n",
    "- **LabelEncoder**: Encode labels with value between 0 and n_classes-1.It can also be used to transform non-numerical labels (as long as they are hashable and comparable) to numerical labels.\n",
    "- OneHotEncoder: Encode categorical integer features as a one-hot numeric array\n",
    "- OrdinalEncoder: Encode categorical features as an integer array."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder, OrdinalEncoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "def label_encode(x):\n",
    "    le = LabelEncoder()\n",
    "    return le.fit_transform(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 0, 0, ..., 2, 2, 0])"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_encode(categ_data[[\"guardian\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Label encoding has the advantage that it is straightforward but it has the disadvantage that the numeric values can be “misinterpreted” by the algorithms. For example, the value of 0 is obviously less than the value of 4 but does that really correspond to the data set in real life?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Activity: Use apply function to encode all categorical data with label encoding"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One hot encoding\n",
    "\n",
    "After label encoding, we might confuse our model into thinking that a column has data with some kind of order or hierarchy, when we clearly don’t have it.To avoid this, we ‘OneHotEncode’ that column.\n",
    "\n",
    "What one hot encoding does is, it takes a column which has categorical data, which has been label encoded, and then splits the column into multiple columns. The numbers are replaced by 1s and 0s, depending on which column has what value.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "def one_hot_encode(x):\n",
    "    one_hot = OneHotEncoder(handle_unknown='ignore')\n",
    "    return one_hot.fit_transform(x).toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 1., 0.],\n",
       "       [1., 0., 0., 0.],\n",
       "       [1., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 1., 0.],\n",
       "       [0., 0., 1., 0.],\n",
       "       [1., 0., 0., 0.]])"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_hot_encode(categ_data[[\"guardian\"]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ord_encode(x):\n",
    "    ord_enc = OrdinalEncoder()\n",
    "    return ord_enc.fit_transform(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       ...,\n",
       "       [2.],\n",
       "       [2.],\n",
       "       [0.]])"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ord_encode(categ_data[[\"guardian\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Question: What is the difference between label, one-hot and ordinal encoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process numerical data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['establishment_year', 'total_toilets']"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_column=list(data.select_dtypes(include=['float64']).columns)\n",
    "total = data[numeric_column].isnull().sum().sort_values(ascending = False)\n",
    "missing_column=pd.concat([total], axis = 1,keys= ['Total']) \n",
    "numeric_column_with_missing = list(missing_column[missing_column['Total']>0].index)\n",
    "numeric_column_with_missing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "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>mathematics_marks</th>\n",
       "      <th>english_marks</th>\n",
       "      <th>science_marks</th>\n",
       "      <th>languages_teacher</th>\n",
       "      <th>school_id</th>\n",
       "      <th>total_students</th>\n",
       "      <th>total_toilets</th>\n",
       "      <th>establishment_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.393</td>\n",
       "      <td>0.338</td>\n",
       "      <td>0.393</td>\n",
       "      <td>2.0</td>\n",
       "      <td>362.0</td>\n",
       "      <td>397.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1950.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.745</td>\n",
       "      <td>0.645</td>\n",
       "      <td>0.745</td>\n",
       "      <td>3.0</td>\n",
       "      <td>357.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1929.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.788</td>\n",
       "      <td>0.655</td>\n",
       "      <td>0.788</td>\n",
       "      <td>9.0</td>\n",
       "      <td>340.0</td>\n",
       "      <td>134.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1976.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.623</td>\n",
       "      <td>0.699</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.0</td>\n",
       "      <td>345.0</td>\n",
       "      <td>143.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1879.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.951</td>\n",
       "      <td>0.704</td>\n",
       "      <td>0.951</td>\n",
       "      <td>4.0</td>\n",
       "      <td>304.0</td>\n",
       "      <td>390.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1914.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mathematics_marks  english_marks  science_marks  languages_teacher  \\\n",
       "0              0.393          0.338          0.393                2.0   \n",
       "1              0.745          0.645          0.745                3.0   \n",
       "2              0.788          0.655          0.788                9.0   \n",
       "3              0.623          0.699          0.623                0.0   \n",
       "4              0.951          0.704          0.951                4.0   \n",
       "\n",
       "   school_id  total_students  total_toilets  establishment_year  \n",
       "0      362.0           397.0            5.0              1950.0  \n",
       "1      357.0            57.0           14.0              1929.0  \n",
       "2      340.0           134.0           15.0              1976.0  \n",
       "3      345.0           143.0           28.0              1879.0  \n",
       "4      304.0           390.0           28.0              1914.0  "
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_data = data[numeric_column]\n",
    "numeric_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "### handle missing values\n",
    "for column in numeric_column_with_missing:\n",
    "    numeric_data[column] = numeric_data[column].transform(lambda x: x.fillna(x.median()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total = numeric_data.isnull().sum().sort_values(ascending = False)\n",
    "missing_column=pd.concat([total], axis = 1,keys= ['Total']) \n",
    "numeric_column_with_missing = list(missing_column[missing_column['Total']>0].index)\n",
    "numeric_column_with_missing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Skealrn offer several method for handling categorical data\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- **MaxAbsScaler**:\tScale each feature by its maximum absolute value.\n",
    "- **MinMaxScaler**: Transforms features by scaling each feature to a given range.\n",
    "- **Normalizer**: Normalize samples individually to unit norm commonly used for text classification or clustering for instance.\n",
    "- **RobustScaler**: Scale features using statistics that are robust to outliers.\n",
    "- **StandardScaler**: Standardize features by removing the mean and scaling to unit \n",
    "- **KBinsDiscretizer**: Bin continuous data into intervals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MaxAbsScaler,MinMaxScaler, RobustScaler, StandardScaler, KBinsDiscretizer "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "def min_max_scaler(x):\n",
    "    transform = MinMaxScaler()\n",
    "    return transform.fit_transform(x)\n",
    "\n",
    "\n",
    "def robust_scaler(x):\n",
    "    transform = RobustScaler()\n",
    "    return transform.fit_transform(x)\n",
    "\n",
    "def standard(x):\n",
    "    transform = StandardScaler()\n",
    "    return transform.fit_transform(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "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>mathematics_marks</th>\n",
       "      <th>english_marks</th>\n",
       "      <th>science_marks</th>\n",
       "      <th>languages_teacher</th>\n",
       "      <th>school_id</th>\n",
       "      <th>total_students</th>\n",
       "      <th>total_toilets</th>\n",
       "      <th>establishment_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.393</td>\n",
       "      <td>0.338</td>\n",
       "      <td>0.393</td>\n",
       "      <td>2.0</td>\n",
       "      <td>362.0</td>\n",
       "      <td>397.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1950.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.745</td>\n",
       "      <td>0.645</td>\n",
       "      <td>0.745</td>\n",
       "      <td>3.0</td>\n",
       "      <td>357.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1929.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.788</td>\n",
       "      <td>0.655</td>\n",
       "      <td>0.788</td>\n",
       "      <td>9.0</td>\n",
       "      <td>340.0</td>\n",
       "      <td>134.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1976.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.623</td>\n",
       "      <td>0.699</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.0</td>\n",
       "      <td>345.0</td>\n",
       "      <td>143.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1879.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.951</td>\n",
       "      <td>0.704</td>\n",
       "      <td>0.951</td>\n",
       "      <td>4.0</td>\n",
       "      <td>304.0</td>\n",
       "      <td>390.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1914.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mathematics_marks  english_marks  science_marks  languages_teacher  \\\n",
       "0              0.393          0.338          0.393                2.0   \n",
       "1              0.745          0.645          0.745                3.0   \n",
       "2              0.788          0.655          0.788                9.0   \n",
       "3              0.623          0.699          0.623                0.0   \n",
       "4              0.951          0.704          0.951                4.0   \n",
       "\n",
       "   school_id  total_students  total_toilets  establishment_year  \n",
       "0      362.0           397.0            5.0              1950.0  \n",
       "1      357.0            57.0           14.0              1929.0  \n",
       "2      340.0           134.0           15.0              1976.0  \n",
       "3      345.0           143.0           28.0              1879.0  \n",
       "4      304.0           390.0           28.0              1914.0  "
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.04040404],\n",
       "       [0.13131313],\n",
       "       [0.14141414],\n",
       "       ...,\n",
       "       [0.3030303 ],\n",
       "       [0.14141414],\n",
       "       [0.01010101]])"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_max=min_max_scaler(numeric_data[['total_toilets']])\n",
    "min_max"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.71428571],\n",
       "       [-0.07142857],\n",
       "       [ 0.        ],\n",
       "       ...,\n",
       "       [ 1.14285714],\n",
       "       [ 0.        ],\n",
       "       [-0.92857143]])"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rbs=robust_scaler(numeric_data[['total_toilets']])\n",
    "rbs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.89838486],\n",
       "       [-0.47187301],\n",
       "       [-0.42448281],\n",
       "       ...,\n",
       "       [ 0.33376048],\n",
       "       [-0.42448281],\n",
       "       [-1.04055547]])"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std=standard(numeric_data[['total_toilets']])\n",
    "std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "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.hist(numeric_data['total_toilets'].values, bins=10);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "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.hist(min_max, bins=10);"
   ]
  },
  {
   "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": [
    "plt.hist(rbs, bins=10);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "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.hist(std, bins=10);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Binning \n",
    "Sklearn provides a KBinsDiscretizer class that can take care of this. The only thing you have to specify are the number of bins (n_bins) for each feature and how to encode these bins (ordinal, onehot or onehot-dense)..The optional strategy parameter can be set to three values:\n",
    "- uniform, where all bins in each feature have identical widths.\n",
    "- quantile (default), where all bins in each feature have the same number of points.\n",
    "- kmeans, where all values in each bin have the same nearest center of a 1D k-means cluster."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bin_data(x):\n",
    "    transform = KBinsDiscretizer(n_bins=100, encode='ordinal', strategy='kmeans')\n",
    "    return transform.fit_transform(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [],
   "source": [
    "bins = bin_data(numeric_data[['total_toilets']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "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.hist(bins, bins=10);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f38d21edf28>"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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(numeric_data[['total_toilets']].values[:50], label=\"orgn\")\n",
    "plt.plot(bins[:50], label=\"bin\")\n",
    "plt.legend()"
   ]
  },
  {
   "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": "3.6.8"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}