3_data_cleaning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Data Cleaning\n",
    "Fix the inconsistencies within the data and handle the missing values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Basic import(s)\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Visualization libraries\n",
    "import missingno as msno\n",
    "%matplotlib inline\n",
    "\n",
    "# Hide all warnings in ipython\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading of the known.csv dataset via pandas.\n",
    "known_dataset = pd.read_csv('data/known.csv', sep=';', # Seperate semi colin in csv file with pandas separator. \n",
    "                            decimal=',', # Specify how decimals interprets.\n",
    "                            encoding='utf-8') \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading of the unknown.csv dataset via pandas.\n",
    "unknown_dataset = pd.read_csv('data/unknown.csv', sep=';', # Seperate semi colin in csv file with pandas separator. \n",
    "                               decimal=',', # Specify how decimals interprets.\n",
    "                            encoding='utf-8') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>index,</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>f1</th>\n",
       "      <th>f2</th>\n",
       "      <th>f3</th>\n",
       "      <th>f4</th>\n",
       "      <th>f5</th>\n",
       "      <th>f6</th>\n",
       "      <th>t1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>14.25</td>\n",
       "      <td>52.17</td>\n",
       "      <td>11</td>\n",
       "      <td>2403.0</td>\n",
       "      <td>890</td>\n",
       "      <td>344</td>\n",
       "      <td>3.0000</td>\n",
       "      <td>497.0</td>\n",
       "      <td>120800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>14.11</td>\n",
       "      <td>52.20</td>\n",
       "      <td>15</td>\n",
       "      <td>5644.0</td>\n",
       "      <td>2659</td>\n",
       "      <td>783</td>\n",
       "      <td>6.7559</td>\n",
       "      <td>757.0</td>\n",
       "      <td>312000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>14.26</td>\n",
       "      <td>52.10</td>\n",
       "      <td>52</td>\n",
       "      <td>2084.0</td>\n",
       "      <td>1438</td>\n",
       "      <td>516</td>\n",
       "      <td>2.3087</td>\n",
       "      <td>550.0</td>\n",
       "      <td>258600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>11.47</td>\n",
       "      <td>53.09</td>\n",
       "      <td>32</td>\n",
       "      <td>3011.0</td>\n",
       "      <td>1287</td>\n",
       "      <td>525</td>\n",
       "      <td>5.0605</td>\n",
       "      <td>529.0</td>\n",
       "      <td>311000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>11.46</td>\n",
       "      <td>53.03</td>\n",
       "      <td>33</td>\n",
       "      <td>2824.0</td>\n",
       "      <td>1797</td>\n",
       "      <td>493</td>\n",
       "      <td>3.6359</td>\n",
       "      <td>523.0</td>\n",
       "      <td>135100</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index,  longitude  latitude  f1      f2    f3   f4      f5     f6      t1\n",
       "0       0      14.25     52.17  11  2403.0   890  344  3.0000  497.0  120800\n",
       "1       1      14.11     52.20  15  5644.0  2659  783  6.7559  757.0  312000\n",
       "2       2      14.26     52.10  52  2084.0  1438  516  2.3087  550.0  258600\n",
       "3       3      11.47     53.09  32  3011.0  1287  525  5.0605  529.0  311000\n",
       "4       4      11.46     53.03  33  2824.0  1797  493  3.6359  523.0  135100"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show head for known_dataset.\n",
    "# We want to make sure that the known.csv dataset has been read in correctly.\n",
    "known_dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>index</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>f1</th>\n",
       "      <th>f2</th>\n",
       "      <th>f3</th>\n",
       "      <th>f4</th>\n",
       "      <th>f5</th>\n",
       "      <th>f6</th>\n",
       "      <th>t1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>13.28</td>\n",
       "      <td>52.63</td>\n",
       "      <td>34</td>\n",
       "      <td>3850</td>\n",
       "      <td>1619</td>\n",
       "      <td>602</td>\n",
       "      <td>5.0465</td>\n",
       "      <td>608.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>13.31</td>\n",
       "      <td>52.45</td>\n",
       "      <td>21</td>\n",
       "      <td>5041</td>\n",
       "      <td>2719</td>\n",
       "      <td>1420</td>\n",
       "      <td>3.5335</td>\n",
       "      <td>1491.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>13.39</td>\n",
       "      <td>52.40</td>\n",
       "      <td>52</td>\n",
       "      <td>1509</td>\n",
       "      <td>674</td>\n",
       "      <td>244</td>\n",
       "      <td>4.9306</td>\n",
       "      <td>225.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>13.46</td>\n",
       "      <td>52.59</td>\n",
       "      <td>42</td>\n",
       "      <td>1291</td>\n",
       "      <td>1535</td>\n",
       "      <td>332</td>\n",
       "      <td>1.9083</td>\n",
       "      <td>345.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>13.22</td>\n",
       "      <td>52.51</td>\n",
       "      <td>27</td>\n",
       "      <td>4742</td>\n",
       "      <td>1682</td>\n",
       "      <td>696</td>\n",
       "      <td>6.1940</td>\n",
       "      <td>775.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index  longitude  latitude  f1    f2    f3    f4      f5      f6  t1\n",
       "0      0      13.28     52.63  34  3850  1619   602  5.0465   608.0   0\n",
       "1      1      13.31     52.45  21  5041  2719  1420  3.5335  1491.0   0\n",
       "2      2      13.39     52.40  52  1509   674   244  4.9306   225.0   0\n",
       "3      3      13.46     52.59  42  1291  1535   332  1.9083   345.0   0\n",
       "4      4      13.22     52.51  27  4742  1682   696  6.1940   775.0   0"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show head for unknown_dataset.\n",
    "# We want to make sure that the unknown.csv dataset has been read in correctly.\n",
    "unknown_dataset.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preparation And Quality Check\n",
    "After we have successfully loaded the data, we need a good overview of quality of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop index columns because it's not necessary.\n",
    "known_dataset = known_dataset.drop(labels='index,', axis = 1)\n",
    "unknown_dataset = unknown_dataset.drop(labels='index', axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import readed .csv dataset into DataFrame.\n",
    "known_dataset=pd.DataFrame(known_dataset)\n",
    "unknown_dataset=pd.DataFrame(unknown_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create column named \"id\" based on row index start with 1.\n",
    "known_dataset['id']=known_dataset.index+1\n",
    "unknown_dataset['id']=unknown_dataset.index+1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>f1</th>\n",
       "      <th>f2</th>\n",
       "      <th>f3</th>\n",
       "      <th>f4</th>\n",
       "      <th>f5</th>\n",
       "      <th>f6</th>\n",
       "      <th>t1</th>\n",
       "      <th>id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14.25</td>\n",
       "      <td>52.17</td>\n",
       "      <td>11</td>\n",
       "      <td>2403.0</td>\n",
       "      <td>890</td>\n",
       "      <td>344</td>\n",
       "      <td>3.0000</td>\n",
       "      <td>497.0</td>\n",
       "      <td>120800</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>14.11</td>\n",
       "      <td>52.20</td>\n",
       "      <td>15</td>\n",
       "      <td>5644.0</td>\n",
       "      <td>2659</td>\n",
       "      <td>783</td>\n",
       "      <td>6.7559</td>\n",
       "      <td>757.0</td>\n",
       "      <td>312000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14.26</td>\n",
       "      <td>52.10</td>\n",
       "      <td>52</td>\n",
       "      <td>2084.0</td>\n",
       "      <td>1438</td>\n",
       "      <td>516</td>\n",
       "      <td>2.3087</td>\n",
       "      <td>550.0</td>\n",
       "      <td>258600</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11.47</td>\n",
       "      <td>53.09</td>\n",
       "      <td>32</td>\n",
       "      <td>3011.0</td>\n",
       "      <td>1287</td>\n",
       "      <td>525</td>\n",
       "      <td>5.0605</td>\n",
       "      <td>529.0</td>\n",
       "      <td>311000</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11.46</td>\n",
       "      <td>53.03</td>\n",
       "      <td>33</td>\n",
       "      <td>2824.0</td>\n",
       "      <td>1797</td>\n",
       "      <td>493</td>\n",
       "      <td>3.6359</td>\n",
       "      <td>523.0</td>\n",
       "      <td>135100</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   longitude  latitude  f1      f2    f3   f4      f5     f6      t1  id\n",
       "0      14.25     52.17  11  2403.0   890  344  3.0000  497.0  120800   1\n",
       "1      14.11     52.20  15  5644.0  2659  783  6.7559  757.0  312000   2\n",
       "2      14.26     52.10  52  2084.0  1438  516  2.3087  550.0  258600   3\n",
       "3      11.47     53.09  32  3011.0  1287  525  5.0605  529.0  311000   4\n",
       "4      11.46     53.03  33  2824.0  1797  493  3.6359  523.0  135100   5"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check whether the adjustments produce the desired result for known.csv dataset.\n",
    "known_dataset.head()"
   ]
  },
  {
   "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>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>f1</th>\n",
       "      <th>f2</th>\n",
       "      <th>f3</th>\n",
       "      <th>f4</th>\n",
       "      <th>f5</th>\n",
       "      <th>f6</th>\n",
       "      <th>t1</th>\n",
       "      <th>id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>13.28</td>\n",
       "      <td>52.63</td>\n",
       "      <td>34</td>\n",
       "      <td>3850</td>\n",
       "      <td>1619</td>\n",
       "      <td>602</td>\n",
       "      <td>5.0465</td>\n",
       "      <td>608.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13.31</td>\n",
       "      <td>52.45</td>\n",
       "      <td>21</td>\n",
       "      <td>5041</td>\n",
       "      <td>2719</td>\n",
       "      <td>1420</td>\n",
       "      <td>3.5335</td>\n",
       "      <td>1491.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>13.39</td>\n",
       "      <td>52.40</td>\n",
       "      <td>52</td>\n",
       "      <td>1509</td>\n",
       "      <td>674</td>\n",
       "      <td>244</td>\n",
       "      <td>4.9306</td>\n",
       "      <td>225.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13.46</td>\n",
       "      <td>52.59</td>\n",
       "      <td>42</td>\n",
       "      <td>1291</td>\n",
       "      <td>1535</td>\n",
       "      <td>332</td>\n",
       "      <td>1.9083</td>\n",
       "      <td>345.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>13.22</td>\n",
       "      <td>52.51</td>\n",
       "      <td>27</td>\n",
       "      <td>4742</td>\n",
       "      <td>1682</td>\n",
       "      <td>696</td>\n",
       "      <td>6.1940</td>\n",
       "      <td>775.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   longitude  latitude  f1    f2    f3    f4      f5      f6  t1  id\n",
       "0      13.28     52.63  34  3850  1619   602  5.0465   608.0   0   1\n",
       "1      13.31     52.45  21  5041  2719  1420  3.5335  1491.0   0   2\n",
       "2      13.39     52.40  52  1509   674   244  4.9306   225.0   0   3\n",
       "3      13.46     52.59  42  1291  1535   332  1.9083   345.0   0   4\n",
       "4      13.22     52.51  27  4742  1682   696  6.1940   775.0   0   5"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check whether the adjustments produce the desired result for unknown.csv dataset.\n",
    "unknown_dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "500001    700\n",
       "162500     86\n",
       "137500     84\n",
       "112500     76\n",
       "187500     75\n",
       "         ... \n",
       "393700      1\n",
       "57800       1\n",
       "360900      1\n",
       "41400       1\n",
       "303100      1\n",
       "Name: t1, Length: 3599, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's see how many target values are in our known.csv dataset.\n",
    "known_dataset['t1'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    767\n",
       "Name: t1, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's see how many target values are in our unknown.csv dataset.\n",
    "unknown_dataset['t1'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1. How many data samples are in our datasets?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The known.csv dataset has 15000 rows and 10 columns.\n"
     ]
    }
   ],
   "source": [
    "# How many datas and columns are in our known.csv dataset?\n",
    "print(\"The known.csv dataset has {} rows and {} columns.\".format(known_dataset.shape[0], known_dataset.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each row cointains the details of the known.csv dataset:\n",
    "15000 data samples are included in our known.csv dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The unknown.csv dataset has 767 rows and 10 columns.\n"
     ]
    }
   ],
   "source": [
    "# How many datas and columns are in our known.csv dataset?\n",
    "print(\"The unknown.csv dataset has {} rows and {} columns.\".format(unknown_dataset.shape[0], unknown_dataset.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each row cointains the details of the unknown.csv dataset:\n",
    "767 data samples are included in our unknown.csv dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2. What format is the data in (int=number, float= decimal number, object=text)? Which formats are surprising?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 15000 entries, 0 to 14999\n",
      "Data columns (total 10 columns):\n",
      " #   Column     Non-Null Count  Dtype  \n",
      "---  ------     --------------  -----  \n",
      " 0   longitude  15000 non-null  float64\n",
      " 1   latitude   15000 non-null  float64\n",
      " 2   f1         15000 non-null  int64  \n",
      " 3   f2         14999 non-null  float64\n",
      " 4   f3         15000 non-null  int64  \n",
      " 5   f4         15000 non-null  int64  \n",
      " 6   f5         15000 non-null  float64\n",
      " 7   f6         14854 non-null  float64\n",
      " 8   t1         15000 non-null  int64  \n",
      " 9   id         15000 non-null  int64  \n",
      "dtypes: float64(5), int64(5)\n",
      "memory usage: 1.1 MB\n"
     ]
    }
   ],
   "source": [
    "# Check data formats from known.csv dataset.\n",
    "known_dataset.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our known.csv dataset cointains 5 columns with integers (int) and 5 columns with float."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 767 entries, 0 to 766\n",
      "Data columns (total 10 columns):\n",
      " #   Column     Non-Null Count  Dtype  \n",
      "---  ------     --------------  -----  \n",
      " 0   longitude  767 non-null    float64\n",
      " 1   latitude   767 non-null    float64\n",
      " 2   f1         767 non-null    int64  \n",
      " 3   f2         767 non-null    int64  \n",
      " 4   f3         767 non-null    int64  \n",
      " 5   f4         767 non-null    int64  \n",
      " 6   f5         767 non-null    float64\n",
      " 7   f6         758 non-null    float64\n",
      " 8   t1         767 non-null    int64  \n",
      " 9   id         767 non-null    int64  \n",
      "dtypes: float64(4), int64(6)\n",
      "memory usage: 60.0 KB\n"
     ]
    }
   ],
   "source": [
    "# Check data formats from unknown.csv dataset.\n",
    "unknown_dataset.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our unknown.csv dataset cointains 6 columns with integers (int) and 4 columns with float."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3. Are there any missing values?\n",
    "\n",
    "In general missing values can be handled by deleting the rows or columns having null values. If columns have more than half of the rows as null then the entire column can be dropped. The rows which are having one or more column values as null can also be dropped.\n",
    "\n",
    "We can handle missing values in the dataset in 7 ways:\n",
    "1. **Deleting rows with missing values:**\n",
    "\n",
    "Pros:\n",
    "- A model trained with the removal of missing values creates a robust model.\n",
    "\n",
    "Cons:\n",
    "- Loss of a lot of information.\n",
    "- Works poorly if the percentage of missing values is excessive in comparison to the complete dataset.\n",
    "\n",
    "2. **Impute missing values for continous variable:**\n",
    "\n",
    "Pros:\n",
    "- Prevent data loss which results in deletion of rows or columns.\n",
    "- Works well with  a small dataset and is easy to implement.\n",
    "\n",
    "Cons:\n",
    "- Works only with numerical continuous variables.\n",
    "- Can cause data leakage.\n",
    "- Do not factor the covariance between features.\n",
    "\n",
    "3. **Impute missing values for categorical variable:**\n",
    "\n",
    "Pros:\n",
    "- Prevent data loss which results in deletion of rows or columns.\n",
    "- Works well with a small dataset and is easy to implement.\n",
    "\n",
    "Cons:\n",
    "- Works only with categorical variables.\n",
    "- Addition of few features to the model while encoding, which maay result in poor performance.\n",
    "\n",
    "4. **Other Imputation Methods:**\n",
    "\n",
    "Depending on the nature of the data or data type, some other imputation methods may be more appropriate to impute missing values.\n",
    "\n",
    "For example, for the data variable having longitudinal behavior, it might make sense to use the last valid observation to fill the missing value. This is known as the Last observation carried forward (LOCF) method.\n",
    "\n",
    "For the time-series dataset variable, it makes sense to use the interpolation of the variable before and after a timestamp for a missing value.\n",
    "\n",
    "5. **Using Algorithms that support missing values:**\n",
    "\n",
    "Pros:\n",
    "- No need to handle missing values in each column as ML algorithms will handle them efficiently.\n",
    "\n",
    "Cons:\n",
    "- No implementation of these ML algorithms in scikit-learn library.\n",
    "\n",
    "6. **Prediction of missing values:**\n",
    "\n",
    "Pros:\n",
    "- Give a better result than earlier methods.\n",
    "- Takes into account the covariance between the missing value column and other columns.\n",
    "\n",
    "Cons:\n",
    "- Considered only as a proxy for the true values.\n",
    "\n",
    "7. **Imputation using Deep Learning Library - Datawig:**\n",
    "\n",
    "Pros:\n",
    "- Quite accurate compared to other methods.\n",
    "- It supports CPUs and GPUs.\n",
    "\n",
    "Cons:\n",
    "- Can be quite slow with large datasets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABdoAAAKACAYAAACLwMohAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAABVuElEQVR4nO3deZhkZXk/7s87M8CwCKISiBoDGhHFZVxwAdGIIjSbKKARA2qUOKwB9YvGaMQlGk3yk4iIC6JiREXEqIPgOiMmiIg6IrixS0ARVFCUBpl+f3+c02NXMyB6znRVd9/3dfVVVedUPTzDmZ6u/tR7nlNqrQEAAAAAAP40C4bdAAAAAAAAzGaCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQzrxWSlnj90Appcx0LwAAAADA7FRqrcPuAYailLKo1npbKWXdJFsm2SzJ5bXWa9r9C2qtE8PsEQAAAAAYfYJ25qVSysJa66pSyt2SnJlkiyT3T3JJki/XWpe2zyvVNwkAAAAAcCeMjmFeakP29ZN8NcktSY5Osn2S05L8fSnl9PZ5QnYAAAAA4E4tGnYDMJOmrVB/bpL1k7wsyfdqrbeWUh6WZFWSb9zJ6wAAAAAAVrOinTmvlPLAUsqjk2aF+pQLnT6s3bSyDdn3T/LuJK+utb65lLJJKWWfydcNp3sAAAAAYNQJ2pmzSmOTJBckeXsp5THJQGh+Q5KN2ufum+S/kvxTrfUtpZSFSfZPslspZfMZbx4AAAAAmDUE7cxZtXFjkgOTbJfkjaWUx055yqVJ7lVK+WiSjyR5Za31ze2+ByV5XpJfJ/nZDLYNAAAAAMwyxUQM5rJSyrrtWJg9k3wqyaeT/Fut9X/b/R9KE6ifkeTvaq3XlVIen+RtSUqSJ9ZabzOjHQAAAAC4Iy6GypxVSlnUhuwbJfl+kv9OsleSWkpJG7a/oH36vkm+1I5vX5Dkl0l2akP2hbXWVTP+BwAAAAAAZgUr2pmTJlegl1LuluRrSa5L8vMkmyR5apLPJTmm1vr19vl/m+QvkmyYZGWST9ZaV7Vh/W3D+DPwxyulLKi1TqxhuzMSAAAAAFhrBO3MWaWURUk+nuR+Sfavtf6wlLJOkmcn+VCSzyd5w+QYmTW83kr2WWTq8WovYHuPNHP4V7UfmqwxhAcAAACAroyOYS7bOMlDkpxRa/1hu22i1vrh0syIOTnJb6eMkRkgZJ89poXs70nyxCTbJPlBkrNKKa+rtd5oZTsAAAAAa8OCYTcAfWnD86kWJJlIssGUbROllIVpVrp/JcnuSd5aSnnIzHTJ2jAlZP+vJE9P8q4keyRZkeTAJMtKKYuF7AAAAACsDYJ25oR2LEgtpSxqx8Ok1np9kguT7FlKeVC7rbah7G1JbkhyRlviB0Nom46mfrhSStkpzUr2w5O8t9b62STvTDOX/wd39DoAAAAA6ErQzpxQa50opayX5H+THFxKWb/ddXSSW5KcXEp58JSXPDjJZkleWWvdoX2974dZoJSyeOoHJ1NC879Kc/bCObXWm9vjvSLJ6Un+odY6Xkp5WillHSvbAQAAAOiTYJG5ZMMkP03ypiTPK6WsW2u9PMkRSe6V5MullJNKKccl+ViSxWkulpmkCeuH0DN/hPbDkC8m+UgpZUnShO3t7tuSrJvktlLKVmk+dPlikhfVWn9bStk3yfOS3GfGGwcAAABgThO0M2fUWn+R5KAkpyV5R5Lntxe/XJZkpzSrm7dL8pQkFyR5fK11VTuznVmg/TDkX5NsmWa2/iOn7L4iyd2SvD7JN9KE7H9Xa72plHKvJM9MM0bmlzPZMwAAAABzXzFBgdmolLKwDckXJIOr0Uspmyd5a5K/SXJYkpNrrbe0+zZNc4HUX03OdK+13jbzfwL+FJPHq5Syc5oL2p6T5Jha63nt/rcl+Yck30nyzFrrFaWUByZ5ZZqLo/51rfX7Q2ofAAAAgDlq0bAbgD9FG7JvkOTkJCeWUj4/GbbXWq8tpRydZJ0kxya5tZTyiVrrTbXW1auZ2wuoCtlniWkfilyT5Lgk/5TkN6WUt9Raz0+zmn1B2guillIWpfl78JdJdhGyAwAAALA2WNHOrDI1bG3ncF+Y5OIkRyX5ymTY3l4g83FJlqVZwf76JO8SrM9O7Qig2t7/SJqL2V6VZoTMNkm+nOQVtdaV7XOen+SRSf48ydlJzqy1XjbznQMAAAAwHwjaGXmllHsk2bTWemn7eOMk+9Ra319K2SbJGUluTnPR06/UWle1z7tbkrOS3DPJtWnGhvgLP4uVUt6U5EVJnpvk3DQXwN05ybvbx6+stX5zeB0CAAAAMB+5GCojrb1Q6VOTLCulPKWUsl6SHyX5u1LKRrXWHyTZPckGSd6e5kKnkx6W5Lr8fjZ3bVe6Mwu1fxeekOR/kiyvtf621npdrfWUJAck2THJq0sp2w2zTwAAAADmH0E7I61dnf7DJJenGQNzRZJvJ/mbWutN7XN+kGS3JIvTzGt/XSnlRUn+I8187svakH2BFe2zx9QPRdqL3q6b5gOV0h7PhVOe94Ukn0vyjCRvKaU8fAgtAwAAADBPCdoZSaWUR5VS9kiSWusFaWasr59mDMxHa61XT31+G7Zvn2Zu9+FJ/iXJr5LsXWudaEP2iZn8M/Cnm/6hSK11otZ6c5JzkuxcStmmvSDuOrXxmyQ/TTOP/QFJbhxO5wAAAADMR4J2Rkpp3CPJ+9KMfpm0RZJPJfnfJO+cDOGnvG5BrfVnaUbH7JhmbvdYrfV37QVUheyzyJSL2p5USvl/U3a9Pc1ZDZ8updyv1vq79nlbpPk78vYkf1VrvXKGWwYAAABgHnMxVEZSKeVhtdbvllLWT/KIWuu57Xz2h6RZrf7kNONjPjPlNZvUWm+cVsdK9lmqlHKvJCck2SfJobXWE9pxMc9M8qYkd0/yziSrkjwyyQ5Jdqi1XjKcjgEAAACYrwTtjJRSSpkcGdLO5f5okicleXGtdVm7/VFJ3pgmbN83yVn5/Yr3t9da/2sYvdO/UspWSV6Z5KAkR9Ra31FKWZTmbIf/l2Sn9qk/TvN35ILhdAoAAADAfCZoZ6RMX4FeSvmbJK9IUpL8c6310+32R6YJ23dOclqale4bJ3nQ5DgRZo9SysL2wrdr2rdVklcleVHasH3KvvsnuSnJrbXWG2aiVwAAAACYTtDOyJhczV5KWTfJRrXWX7Tb905zMdSa5DVTwvZtkhya5IlJfpDkgFrrbe1M9tuG8oegk1LKW5J8ttb6lWnbt0rymiQvSPL3tdYT2+2rz4AAAAAAgGERtDMSJleyt+NivpbkR0mOrrX+pN2/d9YQtrf77pXk521IL2SfRUop60y5oOm9k6xMcmOaD03OnfbcxyT5bJJNk7ys1vr2GW4XAAAAANZowbAbgDYcnyilLE6yfZJfJnlekpeVUrZIklrrfyf55zQjZF5fStlj8vW11uvbkL0I2UdfKWVhKeV+STIlZH9OrfWaJE9L8usk/1VKecLU19Vaz09yfpKrkry2lLLpzHYOAAAAAGsmaGeo2pXst5VS7pbkq2lWra9KclmSl6YJVDdPVoftr0mzqv29pZTtp9YyQmTW2DHJJ0opz0qSUsoZSU4upfxZezHTF6aZu/6hqce4lPJXSW5LcnCSrWutv5z51gEAAADg9hYNuwHmt3Yl+zpJPpnk1jQXu7ywlPKAJM9OE7ynlPK6WutPa62fKqWsn2S3JF8fWuN08Ysk1yU5rZRyfpJ7ppmzf12S1Fq/U0p5QZKTknyslPLGJNcm2T3JVkm+XWv9+TAaBwAAAIA1MaOdoWvHw5yf5Pha65unbN8oyT8keUOStyX598mZ7VOes7DWumom++WPV0rZMMmOtdaz2scPSXMGw6ZJXlVr/dd2+4Ja60R7f6sk706yU5oV7tcn2afW+p0h/BEAAAAA4A5Z0c5QlVJKko3TBK43tdvWq7XeUmu9qZRyapK/T3JUmvnsL22fs6DWOiFkH33tBW4/n+T7pZTltdZbkmye5PtpjumbSimX1VpPTVInPzyptV6e5OmllKcl+W2SK9o57gAAAAAwUqxoZ0a1Fyyta7j/+SRbJtm21vq7ybC93XdWmhD+WUn+vtZ64nC6509VSnlEkstqrb8upTy0HQ+0fpKHpbnI7W5Jnltr/diU16xfa715SC0DAAAAwF3mYqjMmFLKolprLaUsbDctbrcvSPKOJHdPcmYpZd0pIfuSJHdL8p4kZyZ5fill43YlPLNAe/bBd9qQ/a1JPldKeVat9eZa63lp5vCfkeQjpZT92tdskOTdpZRXD7F1AAAAALhLjI5hRrTjQG5r566/q5RyvyS/KaW8tda6vJTyhST/X5rRMN8rpbwryYZJ9k5yY63186WU5yfZLMmvq1MxZoX2rIWJKZvOTLJfkle2AfxptdbzSilvaPd/rJSyU5J7JdkryY4z3DIAAAAA/NGMjmHGtKNCzksynuSnaUfFJHlxrfWkdv9Tkxya5IlJbkjyrSTPTrIwzZzvS9PMbL9V2D6a2jMUdk9yVa11Zbvtg0lOq7V+ppTyxCQfSvLzJP9aaz2tfc6jkhyUZNckP0vz9+K7Q/gjAADMe1PHPDI/OOYA0I2gnbVq8sKW7f39krwwyWG11stKKdsmeWWS/ZMcXGt9z5TXPSDJdbXWX5VS7pHkP9Ksbn9CrfUHM/3n4K5rj9e7kjwmyTOSvCHJTkmelOQ77figHZOcnNuH7eunuTju72qtvxhG/8Bd5xfy+WnyguTD7gNYO0op6ySZmHwPz9zXHvNFro8EAN2Y0c5aVWtdVUrZoJTy4SQ7J7mk1npZu++iJK9L8l9JTiilvGjKSy9vQ/Zdk5zUvvavheyzwg1pwvXfJvlKkick2anWurIN2RfUWr+a5MAk90wzRuZZSdLObb9WyD67tGcxMA+UxnqllHu2H6ROve4Gc1QpZVEp5S9KKY8tpWxQa51wrZS5r5SyUSnlyFLK4mH3wswppWyc5DNpLlTPPNCO9vx0musjbTTsfpgZpZR1SimbD7sPZl77vm6dYfcBc5VwhJmwZZInJ3lxmtXKq9VaL0kTyn4oyTtLKYe32ydXyl2X5ItpQvbvzFTD/OlqrRPtyJefprnA7W8z7XoQ08L2TZL8Wyllj5nulT9dKWXDUspbSyn3bUM3P0/muPaX73ekGeP1v0k+2oauqxz/uas97h/N74/7eaWULSc/OB1ud6wt7Qdo56a5fs6b/UI+P7Qh+wVJNkjytSG3wwxo/40/P8179XOS3DbcjpgJ7XFfkeT/lVLuPeR2mEGllHXTfK8fVkpZb9j9wFxkdAxrxdRxAu0v4o9KE9DcP83FMM+eOm6gHRXztiSbJnnStH2rx88wO7QjYJ6dZFWSf0wzY/+FSc5tw5mSJO39pyQ5NsnetdbLh9Qyf4RSygZpzlZ4dJKPJzmy1voT4yTmrvYXsq8luSnNtTY2SzKW5roZO9Rabxlie6wl7XH/RprrZnw8yTpprqPyszQ/qwUyc1gp5fQkW6dZMHFqkr93zOeuNmT/TpLLkhxYa716Dc/xc34OaT9Qe3+Se6dZEPVjx3fua8PVT6Q5a2U8ze/gx9VafzrUxpgRpZRNk7wvyS5JjkryQe/joV+Cdno1PRSffENeSlmUZEmaudwLkrwgydenBer3SfKTyVPSzf2dPe7oF6/2Q5ZHpjnuC5O8sNb6tXbfhkm2qbV+s5SyuNY6PqNN8ydpj+mxSZ6Z5KIkD0sTwB5Ra73GL+FzT7uS9VNpQtaDaq1XtNuPSPLmJC+otX7cv9tzS/uL+OeSTCR50eQHoaWUf0zzwek2vtfnplLKOrXW35VS3p3kQUk+luQ/k3wwySG11t8NtUF61y6QuDDJFWmuiTTe/h34yyT3SvJnad63G+03h5RS7pZkeZozi4+vtd5WStk+yfZJHpjmrOKVtdaLh9gmPWoXOx2Z5KVJ3p5m4cTLk/xrkrcL2+e+9u/APZP8W5K/TbOA4r9qrb8damMwhzjll96UUhZNmcl+RCnlP5P8eynlMbXW22qt5yc5IElN8oEkj5s647XWevXkCAphzezRfrgy0d7frpSyTynlcaWUe7bbv5Pk+WlWt7+vlPLk9he3Y5OcVkq5u5B9dmg/HNkxyTZpTjd9RpqVUNsneXsp5c+NkZlb2hXNf5tkiyT/ORmytz6YZoX7Y5LmDJUZb5C1oj3uS5PcnORfaq2Xtx+YJ81q12uSHF1KObaU8oT2LBdmuXYk2L2mBOmfTHJrmg/aXpNmkcQJU56/7Yw3Se/a7/dXJblHmu/537Qh+7OSLEvy5TQz2y8opRxYSrn70JqlF+33+uZJFid5QJozjW8rpfxNki8keUmSvZKckuQ9bfjOHNC+V7t7kt8leVet9egkb03yyiRHlFK2mHyua7HMPe0H6bXWen2aD1e+mOb4P9v1WKA/VrTTiykr1++WZnXrojRv3hanWQnzhiTvrLVeV0p5TJqgvSQ5uNZ69pDapqOpZzCUUj6QJnC9b5IfJ7k8ydJa65VtQPPIJCemWQH9wzSro3attX5jGL3zx2m/t3+U5H+SfCvJie33c0nzRu2ANPP+Jle2LzJiYHZrj/llSX6QZmbrPpOrGad8mLI8ybW11mc75nNDe9wvSrIyydlpVjne3O7bIMk3k9wtya/TnKl0nzSr405MfOAyW035fv90khe3o90enuSraX62X5HksDTv505K897uL9L8HP/lUJqms/a4fzfJt9P8W39QkrPSjAr6aJqgdUWanwF/l+YC9wfVWj/qDLbZacr3+ulpvqcvTPNBynuTnJHm+/tDtdar2jPXXpbm78bf11qvHE7X9GHq92wpZeNa668mf5crpbwpTdhuZfscNe34n5Dkr9KsbF+S5ppqL40xMtCLRX/4KXDn2nEBE+2FNT6R5OdJDk/zpuz+aVbFvTbNSvbXp3kz//w0p6QfnOYXeWaZaSH7B9OsdP6HNKufPpVkjySnl1L2q7VeVkr5VpKnpjk9cSLJB2qtPxpO9/wx2g9KTk0zu3O/qfvaUO0V7aKXA9KsbP+HWuvVbQi/Ufu0m2a6b/50U475ZbXWHUsp69Zab53cP+WN+jVpArek+Td+4BodzC5TjvtPaq17Tdu3IMnX06xw3iNNULNOmoD99Uk+VWv92cx2TB+mHPeLkxzehuwb1VovKKWsTPJXtdaLSinvSrMK8q3t7bOE7LPXlON+ba31me1qxvWT7Jvmekr/kuQ/aq2/aZ9/Zpr3d/9SSjmj1vrrIbXOn2ja9/rL2jMXfpjksUmuTPMB6sfTnLWUWuvb23/735pkq/Y5zDKT78umfTA2eb2sVe3tq9r38q9sX/K2WuvPSnMdtT+bHP3J7DSZ17T3T0yya5rf27+V5BFJDkxyXLv/ZGebQzdO7+dP0o4IOSAZWL32l0kenGYVxAVtKPPDNJ+OHpfkNaWUR7U/0L+V5IlpRhIwS5RSNiql7JX8/o1ZKeX5SbZNM3/9M2lWNu6WZkXMFkk+Vkq5X611Va31+lrrK5O8Rsg+q/x5kvsl+e8kKaW8r5TyilLKgtJcSCu11lekmfG5fZL/LKX8eZoA9tgkzzVOZtYZOOZpTh0/eg3HcWGaM5fSroi6W5LXllK2mbFO6dP07/WTJr/X2/2vSrJ7rXVlrfVXtdafpxklsnmaoIbZafK4f6bW+tv2l/BD2n3Xp5nJn1rrjUmelObieYuS7O3f9llt+r/zb04zOuYj7dcpU0L2he3x/1iaC2c+ZMa7pQ9Tv9dvKqW8Jc33+CPTrGTeuNZ6cfvzfHKMxH8muSXN+ztmmXbRy76llNdN2XZmmnGeC6c+t9b6qiRvSfKKJP9QSnlcmuP/P6WUTYySmV1KKRuXUvZLfp/XlFLul2TnNGcrfqLWenmt9b+THJFmweSxSfY3EhC6saKdP1q7GuLItOFZrfWD7a57pzmF/Mb2eevWWm9tfya/O8lz08z/+nb7j/0P2ucNXECVkfbaJC8rpbyw1vrB9g3X4iRfqLV+pZRyUJI3JnlOrfUTpZRj0/zg/mgp5cBa6yXJ70N6Zo2fJLkuyUGllKelGf/zlCkrIxa2H6S8opRS06yKeG+aEHaXJP+f08tnnTUd839bw3H8bdr3EqWUTdL8ov6SNGEMs8+ajvt/TDnun1nDa+6dZlTYhTPTImvBmo7709p9X06yZ5KUUk5Ns0ji+WlWt741Tej+DzPdML2YftyXpAlcb0yy9eR7tnYl5OT7tg3TBLPGSsxO04/5I5M8Ks3IqHcn2ao0F7z+t1rreBvEbpvkF2kWTzH7rJtmMdz/K6XcN82okIek+V3tdr+P1Vr/sX0v//I07+c3TLJd+0Ebs8u/JllSSvnv+vtrr2yQZnHEr5Pf5zVJrk7yzjRnLP5rkg1KKe+ZejYrcNeZ0c6fpF2t+rYkz07yolrr+0spm6b5Rft/aq3PaZ83GbZvkOYN2intyldmoWnH/cW11pPaD17uneRXaWa5fjzNG/Sb20/Nz0kzPuQ7aUbHrDJWYvaYPN20lLJVmjmuC9NcW+ED0563ej53KeWdaUZG3ZAmkP/OzHZNF3f1mLfPfXuSJ6cZHXVskuckeWKt9dsz1zF9+CO+16fO+Nw8zXHfJMnf1Fp/NbNd09UfOu6llEekmdn9szQz2Z9Ta/1C+57vgDQftH9/ON3zp7qD435Imvm8E1OeN/Vn++ZpRkUtSvP3wPf7LHJHx7z9HW5hkmelCdpKkren+SDtIWk+PB9L87P9x8Ppni7ahRCvTrNQ7uY0x/KCcgfXWWjHxSxPE7I/udbqg/RZqP05fUt7ptoOtdb/LaWsk+Y6PN+vtT6jfd7q8ZCllP9Nc4HkBWk+cL1hSO3DrOZ0T/4ktdafpFnB9Ik0p569uJ3T+e4k+5XmgiqZ8inoNmlWPV0+jH7px7TjfmJ73G9r33jfP80Kt5VtyF7SjBK4Ksk/pRktc5uQfXaZcrxenObDlGuSvKqU8tipp5DWWm9rR8ncJ83IoBvSvJEXss8yd+WYTzn2pf36lzRnLQnZZ6m7etynhOwPSDNq4ilJXi50m53u5Lg/rg3ffpbmDMSS5G+SfKl93S+TvEPIPjvdwXH/xySPmfb9PhmyPyjNSIkd4/t9Vrqzf+PbsxI/nmZu8zVpRof8Jskn04TsewnZZ68pq9FvSvNv+Uvb7RPtgqnVSil/meTf0ox/FLLPYrXWX7Yh+8FJvlpKOahd2f62JHuWUt7cPm8yZH9Qmg9i9kvyECE7/OmsaKeTdnXLO5LskyZk+USS9yd5XpLT2sd3T/J3aVZOPM7YkNlv2nGfPKPh3km+n+ashYPbN2r/mOYUtRc79Wx2a4/n4iR3S/M9vl6SA2ut5055zvppxgsdneRRtdaVQ2iVntzFY/5vSV6W5rTyp9davzWMXunPXTzur06yd5J7JHmmD9Rmvzs47gfUWr9eSnlUmosdf3cyeGVuuIvf769Ncybj+vH9Puut4ZgvTnN2yjdqM5v9HkkelGaU0JVpvu+vGlK79KQ97n+e5gPT5yf5dK31+e2+1WNcSylPTPLBJPt4Hz83tGekvz3N7+1/W2s9pZTyvjR/D05JcnySTdOclfr4JDu1i+uAP5Ggnc7a0PX4NKcc7ptmzt8r0oyOuE+SHye5IM0P7N8VM9nnhGlh+0G11ve1cx3fmCZ0+1WaH9p/XWu9YHid0rdSynZpfjlbN1N+IW9XxTwtyaW11ouH2CI9m37Mk3y9PQV9jzSjQ55Ra71oiC2yFqzpe71d7bpzmpXsJ9ZaLx1mj/RvynFfnOR5tdavD7klZsCd/GzfPs0Kx+NrO7eduWENP9vP92Ha3FZK2SzNmcbTw/b105ztcGKSdauZ7HPKtLxmvzRnq7wyyVFpFk3clOZs5Gf4gAW6E7TTi/Yf73cmeWaS/WutHy2lbJzmAjo/SXJlG8os8gZu7pgWtk8e993T/BD/SZKTa60/GmaP9K8N2h6T3/9y9rdpfjlzwdM5ag3H/IAk57X313d66dy0pu/1Wut57T4/z+eoacd9vTRnKX7D6Le57Q5+tn/D+/e5645+tvten9tKKX+W5FVJXpDmGhxvTnONhoMy5ULIzC13kNf8WZqFE9cl+WGt9eph9ghzhaCd3kz7pPSgWuv7pu1f4wVXmN2mHfcDa63/1W53vOewKb+cvSfNBfJ2rbWeP9yuWJsc8/nJcZ+fHPf5yXGffxzz+akNWF+W5OAkt6WZzb2H6+zMbdN+b39RrfX9Q24J5iQXQ6U3tdZrkxya5ONJ3ltK2W3afqHrHDTluJ+a5ORSyt+22x3vOaxd7XR+mhUwP0ryy+F2xNq2hmN+w1AbYkb4Xp+fHPf5yXGffxzz+anW+rMkb0pzAdzDk2wvZJ/7pv3e/r5Syv5DbgnmJCva6V17wY1DkxzjNNP5oz3ub07y1lrr94bdDzOnlLK41jo+7D6YOY75/OS4z0+O+/zkuM8/jjnMD35vh7VL0M5aZabj/OJCtwAAADC6/N4Oa4+gHQAAAAAAOjCjHQAAAAAAOpiXQXspZd9SynGllK+WUn5VSqmllP8adl8AAAAAAMw+i4bdwJC8OskjktyU5P+SbDPcdgAAAAAAmK3m5Yr2JEcl2TrJxkkOHnIvAAAAAADMYvNyRXutdfnk/VLKMFsBAAAAAGCWm68r2gEAAAAAoBeCdgAAAAAA6GBejo7pw1//9V/XYffAzDr22GOTJEceeeRQ+2BmOe7zk+M+/zjm85PjPj857vOT4z7/OObzk+M+f61YsWIuzoaeE9nj2NhYnvGMZ2Tp0qXDbmWt/x2xoh0AAAAAADqwoh0AAIA5b8mSJVmxYsWw22AGrVy5ctgtAMxqF154YQ4//PDOda655poeuhl9gnYAAADmvJUrVxonMY9MjhAB4E/3ne98p5c6l19+eS91Rp2gHQAAAACAAc973vPyvOc9r1ONsbGx7LDDDj11NNrMaAcAAAAAgA7m5Yr2UsreSfZuH27R3j6hlPKB9v71tdaXz3BbAAAAAAAj4Utf+lLe+MY3dq5z4YUX9tDN6JuXQXuSJUmeP23b/duvJLkyiaAdAAAAAJiXzjjjjF7qXHTRRb3UGXXzcnRMrfWYWmu5k68th90jAAAAAMCwXHvttcNuYVaZl0E7AAAAAAB37M1vfnMvdXbbbbde6ow6QTsAAAAAAAP++7//u5c6X/nKV3qpM+rm64x2AAAAAADuwHOe85xcfvnlqbWmlJIkq28n769p+9R95513XnbaaaeZa3qIBO0AAAAAAAzYfPPN87a3va1TjbGxsWywwQY9dTTajI4BAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHSwadgMAAAAAAIyWVatW5bvf/W6SpNY6cDv9/vRtk7fj4+Nru82RIWgHAAAAAGDA8ccfn09+8pOd65xxxhlZunRpDx2NNkE7AAAAAAADnvOc5+Tyyy9fvTq9lDJwO9XUfVP3n3/++fnrv/7rtd/sCBC0AwAAAAAwYPPNN8/b3va2TjXGxsay4YYb9tTRaBO0AwAAAAAw4JZbbsmZZ56ZWmunGe2rVq2agW6HT9AOAAAAAMCAd7zjHVm2bFnnOsuWLcuhhx7aQ0ejTdAOAAAAAMCAgw46KOuss87tZrRPmvp4+vz2yduPf/zj2W233Wai3aETtAMAAAAAMOB1r3tdvvWtb3Wuc/rpp+fwww/voaPRtmDYDQAAAAAAMFo23njjXupssMEGvdQZdVa0AwAAAAAw4LWvfW1e+9rXdqoxNjaWPffcs6eORpsV7QAAAAAA0IEV7QAAAAAADPjkJz+Zt7/97Z3rfP3rX8/SpUt76Gi0WdEOAAAAAMCALbbYYqTqjDpBOwAAAAAAA84999xe6lxwwQW91Bl1RscAAAAAADDg+c9/fm688cbUWpMkpZSB/aWU1dvuaN8Xv/jF7LzzzjPT8JAJ2gEAAAAAGHCPe9wjxxxzTKca//M//5PFixf309CIMzoGAAAAAAA6ELQDAAAAAEAHgnYAAAAAAOjAjHYAAAAAAAb87ne/yznnnJOJiYnVF0SdNPm41jqwb/r98fHx2712rhK0AwAAAAAw4OCDD86ll17auc6pp56agw8+uIeORpvRMQAAAAAADFiyZEkvde51r3v1UmfUWdEOAAAAAMCAww47LIcddlinGmNjY3nqU5/aU0ejzYp2AAAAAADowIp2AAAAAAAGfP/7388hhxzSuc5Pf/rTHroZfVa0AwAAAAAw4Bvf+EYvdS6++OJe6ow6K9oBAAAAABhw4IEH5sADD/yTXltrTdLMaN9xxx37bGtkCdoBAAAAABhw5ZVX5uijj86qVatWB+eTt5Nqrau/pqu15pZbbskvfvGLGel32ATtAAAAAAAMOPPMM/Ozn/2sc51vfetbPXQz+gTtAAAAAAAMeMlLXpJ99tknpZQkud3tHW1Lfr/y/bnPfW6e9rSnzUS7QydoBwAAAABgwC233JKvfOUrmZiYWL1t+giZO7qdvH/LLbdk1apVM9XyUAnaAQAAAAAYcPzxx2fZsmWd6yxbtiyHHnpoDx2NNkE7AAAAAAADXvKSl2Tx4sWrH/8xI2Qmb0855ZTsvvvuM9LvsAnaAQAAAAAYsNFGG3VeiX766adn0aL5EUEvGHYDAAAAAAAwmwnaAQAAAACgg/mxbh8AAAAAgLvshhtuyLvf/e6sWrVq9bZa6126nbw/Pj6em2++eaZaHipBOwAAAAAAA0466aScddZZnet88YtfzFFHHdVDR6NN0A4AAAAAwIDDDjss22677erHpZSB2zVtm77v9a9/ffbYY4+ZaHfoBO0AAAAAAAxYd911s8suu3Sq8da3vnUgfJ/LXAwVAAAAAAA6ELQDAAAAAEAHgnYAAAAAAOhA0A4AAAAAAB24GCoAAAAAAANuu+22nHvuuZmYmEitdfX2Wuvqr1WrVg08rrVmYmIiSTIxMZHx8fGB185lgnYAAAAAAAbsvPPOvdQ59dRTc/DBB/dSa5QZHQMAAAAAwIB/+Zd/6aXODjvs0EudUWdFOwAAAAAAA7bffvssX768U42xsbHc97737amj0WZFOwAAAAAAdCBoBwAAAACADgTtAAAAAADQgaAdAAAAAAA6ELQDAAAAAEAHi4bdAAAAAAAAo+WMM87Iv//7v3eus3Llyu7NzAJWtAMAAAAAMKCPkD1JfvjDH/ZSZ9QJ2gEAAAAAGHDve9972C3MKoJ2AAAAAAAGPOQhD+mlzgYbbNBLnVEnaAcAAAAAYMBWW23VS5173vOevdQZdS6GCgAAAADAgP333z/7779/pxpjY2PZfvvte+potFnRDgAAAAAAHVjRDgAAAADAgK9//et55Stf2bnOpZde2kM3o8+KdgAAAAAABvzkJz/ppc4NN9zQS51RZ0U7AAAAAAAD9t577+y9996daoyNjeXRj350Pw2NOCvaAQAAAACgA0E7AAAAAAB0YHQMAAAAAAADzjrrrLzlLW/pXOfb3/52D92MPivaAQAAAAAYsP766/dSZ8MNN+ylzqizoh0AAAAAgAFPfvKTs3z58k41xsbGsvXWW/fU0WgTtAMAADDnLVmyJCtWrBh2G8yglStXDrsFgFntW9/6Vl72spd1rvPjH/+4h25Gn6AdAACAOW/lypU58sgjh90GM+TYY48ddgsAs96ll17aS52f/OQnvdQZdYJ2AAAAAAAG7Lffftlvv/061RgbG8vjHve4njoabS6GCgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0sGjYDQAAAAAAMFpuueWWfPazn83ExERqrQP7Jh/f0e3k/fHx8axatWqGOh4uQTsAAAAAAAOOO+64nHHGGZ3rLFu2LIceemgPHY02o2MAAAAAABjQ10r0BQvmRwQ9P/6UAAAAAADcZXvvvXcvdXbYYYde6ow6QTsAAAAAAANe+cpX9lLnC1/4Qi91Rp2gHQAAAACAAQceeGAvdbbZZpte6ow6F0MFAAAAAGDAHnvskXve856ptabWmiS3u528v6btSfKmN70pD3/4w2eo4+EStAMAAAAAMGDXXXfNxMRE5zqnnnpqDj744B46Gm1GxwAAAAAAMOClL31pL3WsaAcAAAAAYF7afffds/vuu3eqMTY2lgc/+ME9dTTarGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANDBomE3AAAAAADA6Kq1rv6a/vgP7ZsvBO0AAAAAAAz4wAc+kA9+8IOd6yxfvjxLly7toaPRZnQMAAAAAAADnvCEJ/zRrymlpJSSBQsWZMGCJnp+6EMf2ndrI8mKdgAAAAAABjzoQQ/K8uXLO9UYGxvLZptt1lNHo82KdgAAAAAA6EDQDgAAAAAAHRgdAwAAAADAgC9/+ct5wxve0LnORRdd1EM3o8+KdgAAAAAABqxatWrYLcwqVrQDAAAAADBg5513zs4779ypxtjYWLbddtueOhptVrQDAAAAAEAHgnYAAAAAAOhA0A4AAAAAAB0I2gEAAAAAoANBOwAAAAAAdCBoBwAAAACADgTtAAAAAADQgaAdAAAAAAA6ELQDAAAAAEAHgnYAAAAAAOhg0bAbAAAAAABgtHzhC1/Im970ps51Lrjggh66GX1WtAMAAAAAMGDhwoW91FlnnXV6qTPqrGgHAAAAAGDATjvtlJ122qlTjbGxsTz4wQ/uqaPRZkU7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOFg27AQAAAAAARss111yTY445JhMTE6m1DuybfFxrHdg3/f74+HhuvPHGmWl4yATtAAAAAAAMOPLII3Pdddd1rnPWWWflFa94RQ8djTZBOwAAAHPekiVLsmLFimG3wQxauXLlsFsAmNUe/OAH9xK0b7zxxj10M/oE7QAAAMx5K1euzJFHHjnsNpghxx577LBbAJj1Hv3oR+fss8/uXGfLLbfs3swsIGgHAAAAAGDAXnvtlb322qtTjbGxsTz4wQ/uqaPRtmDYDQAAAAAAwGxmRTsAAAAAAANuvvnmnHbaaZmYmFi9rdZ6l24n74+Pj+e2226bqZaHStAOAAAAAMCAd77znVm2bFnnOmeccUYOO+ywHjoabYJ2AAAAAAAGLF26NJtuuunqx6WUNd5OVUpZ/ZUk73vf+7LHHnvMQLfDJ2gHAAAAAGDAhhtumL/7u7/rVOPDH/5wFi5c2FNHo83FUAEAAAAAoANBOwAAAAAAdCBoBwAAAACADsxoBwAAAABgwHXXXZf/+I//yMTERGqtA/smH9daB/ZNvz8+Pp6bbrppZhoeMkE7AAAAAAADPvzhD+frX/965zpnn312Xv7yl/fQ0WgTtAMAAAAAMODwww/P9ttvf7vtpZTVt5P372jfUUcdld12223tNzsCBO0AAAAAAAxYuHBhHvvYx3aqsXjx4p66GX0uhgoAAAAAAB0I2gEAAAAAoANBOwAAAAAAdGBGOwAAAAAAA84999z84z/+Y+c6l156aQ/djD4r2gEAAAAAGHDVVVf1UueXv/xlL3VGnaAdAAAAAIABP/7xj3upc/XVV/dSZ9QZHQMAAAAAwIAXv/jFWbBgQWqtKaUM7CulrP6avn3q/tNOOy1jY2Mz1vMwCdoBAAAAABiwySab5KijjupUY9myZVl33XV76mi0GR0DAAAAAAAdCNoBAAAAAKADQTsAAAAAAHQgaAcAAAAAgA4E7QAAAAAA0IGgHQAAAAAAOhC0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdLBp2AwAAAAAAjJbjjjsup59+euc6n/vc57J06dIeOhptVrQDAAAAADCgj5A9SW644YZe6ow6QTsAAAAAAAPe8Y539FJnp5126qXOqDM6BgAAAACAAdtuu22WL19+p8+ptWZiYmL1/cmvycfPeMYzstlmm631XkeBoB0AAAAAgAEf+tCHctJJJ3Wus2LFCjPaAQAAAACYfx796Ef3UuchD3lIL3VGnaAdAAAAAIABF1xwQS91Lrvssl7qjDpBOwAAAAAAAz7ykY/0UufKK6/spc6oE7QDAAAAADDgoIMO6qXO1ltv3UudUSdoBwAAAABgwJlnntlLnR/96Ee91Bl1gnYAAAAAAAbsueeevdRZsmRJL3VGnaAdAAAAAIABl1xySS91rrjiil7qjDpBOwAAAAAAAzbZZJNe6ixevLiXOqNO0A4AAAAAwIBHPepRvdTZZptteqkz6hYNuwEAAAAAAEbLtttum+XLl3eqMTY2ls0337ynjkabFe0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6MDFUAEAAAAA+KPVWtd4O33/fCBoBwAAAABgwAc/+MF84AMf6Fzn7LPPztKlS7s3NOKMjgEAAAAAYMDjH//4XupsvfXWvdQZdVa0AwAAAAAw4EEPelCWL1/eqcbY2Fi22GKLnjoabVa0AwAAAABAB1a0AwAAAADwB9VaV39NPp6YmLjDfS6GCgAAAADAvNXXxVCXL1/uYqgAAAAAAMw/119/fS91br311l7qjDpBOwAAAAAAAx7/+Mf3Uuev/uqveqkz6oyOAQAAAABgwA477JDly5d3qjE2NpYHPOABPXU02qxoBwAAAACADqxoBwAAAABgQK01v/rVr1bfX9Pt5P01bU+SiYmJmWh1JAjaAQAAAAAYcOKJJ+aUU07pXOfzn/98li5d2kNHo03QDgAAAADAgKc97Wn57Gc/u3pVeilljbd3tu/666/PdtttN2M9D5OgHQAAAACAAVtttVU++clPdqoxNjaWTTfdtKeORpuLoQIAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHZjRDgAAAADAgF/84hc57rjjMjExkVrrwL7Jx7XWgX3T74+Pj+fmm2+emYaHTNAOAAAAAMCAD37wg1mxYkXnOl/60pdy1FFHdW9oxAnaAQAAAAAYcNhhh+VRj3rU7baXUgZu13R/8vE//dM/Zffdd1/LnY4GQTsAAAAAAAPWWWedPPnJT+5UY/HixQMh/FzmYqgAAAAAANCBFe0AAAAAAAy45pprcswxx3S+GOqNN944Mw0PmaAdAAAAAIABp59+ei6++OLOdc4999weuhl9gnYAAAAAAAYccsgh2WWXXW43Y33qxVD/0L4XvvCF2WWXXWam4SETtAMAAAAAMOCcc87Ja17zms51rrjiiu7NzAIuhgoAAAAAwIDLLruslzq//OUve6kz6gTtAAAAAAAM+PnPf95LnauvvrqXOqPO6BgAAAAAAAYcfvjheexjH7v68dT569Pd0dz2V7ziFdljjz3WcqejQdAOAAAAAMCARYsW5XGPe1xqram1rt4++Xhy28TExOr70/ett956M9/4kAjaAQAAAAAYcNJJJ+VDH/pQ5zpf+tKXsnTp0h46Gm1mtAMAAAAAMGD77bfvpc6SJUt6qTPqBO0AAAAAAAw488wze6lzzjnn9FJn1AnaAQAAAAAY8NznPreXOjvvvHMvdUadoB0AAAAAgAEf/vCHe6nzta99rZc6o87FUAEAAJjzlixZkhUrVgy7DWbQypUrh90CwKy29dZb91Jn880376XOqBO0AwAAMOetXLkyRx555LDbYIYce+yxw24BYNYbHx/vpc7NN9/cS51RZ3QMAAAAAAADrrjiil7qXH311b3UGXWCdgAAAAAA1opSyrBbmBGCdgAAAAAABixY0E90vGjR/JheLmgHAAAAAGDAHnvs0en1kyvZH/OYx/TRzsgTtAMAAAAAMODaa6/t9Ppaa5Lk5z//eR/tjDxBOwAAAAAAA37zm9/0UufWW2/tpc6omx8DcgAAAAAAuMvGxsYyNjbWucbDH/7wnjoabVa0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdmNEOAAAAAMCAH//4x/mnf/qnrFq1KrXWJFl9O6nWuvprulprxsfH88tf/nJG+h02QTsAAAAAAAPOOOOM/N///V/nOueff34P3Yw+QTsAAAAAAAOWLl2avfbaK6WUlFKSZPXt5P01bZ/6+HnPe1523nnnGep4uATtAAAAAAAMKKXkPve5T6caCxbMn0uECtoBAAAAABhw88035zOf+UwmJiZWb5ucxT65Yv2OHk9uGx8fn6l2h07QDgAAAADAgHe+851ZtmxZ5zpf/vKXs3Tp0h46Gm2CdgAAAAAABrzkJS/Jhhtu2KnGxz72sTzlKU/pqaPRJmgHAAAAAGDARhtt1Hkl+qc+9anbXSh1rhK0AwAAAAAw4OMf/3je+c53dq5zzjnnzIvRMfPnsq8AAAAAANwlD3jAA3qpc7/73a+XOqPOinYAAAAAAAY86lGPyvLlyzvVGBsby33ve9+eOhptVrQDAAAAAEAHVrQDAAAAADDg05/+dN72trd1rvONb3zDjHYAAAAAAOafzTbbrJc697rXvXqpM+qsaAcAAAAAYMATnvCEXma0b7XVVj11NNoE7QAAAAAADDj11FNzwgkndK5zzjnnGB0DAAAAAMD888AHPrCXOve73/16qTPqrGgHAAAAAGDAIx/5yF5Gx9z3vvftqaPRZkU7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6GDRsBsAAAAAAGC0XHPNNTnmmGMyMTGRWuvAvsnHtdaBfdPvj4+P58Ybb5yZhodM0A4AAAAAwIDTTz89F198cec65557bg/djD5BOwAAAAAAAw455JDssssuKaUMbJ98XEr5g/te+MIXZpdddpmZhodM0A4AAAAAwIAFCxbkgQ98YKca6667bk/djD4XQwUAAAAAgA4E7QAAAAAA0IHRMQAAAAAADLjhhhvy7ne/O6tWrVq9rdZ6l24n74+Pj+fmm2+eqZaHStAOAAAAAMCAk046KWeddVbnOl/84hdz1FFH9dDRaBO0AwAAAAAw4LDDDsu22267+nEpZeB2Tdum73v961+fPfbYYybaHTpBOwAAAAAAAy666KL867/+a+c6V111VQ/djD4XQwUAAAAAYMDFF1/cS51rrrmmlzqjzop2AAAAAAAGPPvZz86zn/3sTjXGxsby+Mc/vqeORpugHQAAAACAATfeeGNOOumkrFq1avW2Wutdup28Pz4+nvHx8ZlqeagE7QAAAAAADDjxxBOzbNmyznU+//nP58gjj+ze0IgTtAMAAAAAMOCFL3xhVqxYkYmJiTXuL6X8wRq/+c1vsuuuu/bd2kgStAMAAAAAMOD9739/brrpps51zjzzzBxxxBE9dDTaBO0AAAAAAAw45JBDct/73nf14+kr2Kc+nn5/8vFxxx2XPffccy13OhoE7QAAAAAADPjCF76Qd73rXZ3rfPOb3+yhm9G3YNgNAAAAAAAwWjbbbLORqjPqrGgHAAAAAGDAE57whCxfvrxTjbGxsWy55Zb9NDTirGgHAAAAAIAOrGgHAAAAAGDArbfems997nOptabWmiS3u528v6btSTI+Pp6JiYkZ6ni4BO0AAAAAAAx46UtfmosuuqhznY9//OM55JBDeuhotBkdAwAAAADAgPvf//691Nl00017qTPqrGgHAAAAAGDAUUcdlYMPPjillCRZfTtp6siYqdum3u6zzz55+tOfPgPdDp+gHQAAAACAAaWUrL/++p1rzBdGxwAAAAAAQAdWtAMAAAAAMOD000/Pcccd17nOueeem6VLl/bQ0Wizoh0AAAAAgAF/8Rd/0Uud+9znPr3UGXVWtAMAAAAAMGC77bbL8uXLO9UYGxvrLbAfdVa0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdCNoBAAAAAKADQTsAAAAAAHQgaAcAAAAAgA4E7QAAAAAA0MGiYTcAAAAAAMBoOeuss/KWt7ylc52LLrqoh25GnxXtAAAAAAAM6CNkT5ILL7ywlzqjTtAOAAAAAMCAo446qpc6D3rQg3qpM+qMjgEAAAAAYMBee+2Vvfbaq1ONsbGxLFmypJ+GRpwV7QAAAAAA0IGgHQAAAAAAOhC0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdCNoBAAAAAKADQTsAAAAAAHSwaNgNAAAAAAAwWr7yla/kmGOO6Vzn+9//fvdmZgEr2gEAAAAAGHDzzTf3Uue2227rpc6os6IdAAAAAIABu+66a3bddddONcbGxvKwhz2sp45Gm6AdAAAAAIAB733ve3PKKad0rrN8+fIsXbq0h45Gm9ExAAAAAAAM6CNkT5Kf/exnvdQZdYJ2AAAAAAAGvOENb+ilznbbbddLnVFndAwAAAAAAAOe+MQn5otf/GKSpNa6+uuPefzsZz8797///YfzB5hhgnYAAAAAAAacdNJJ+dCHPtS5zpe+9CUz2gEAAAAAmH923HHHLFq0KAsWLMiCBQuycOHCLFy4MIsWLcqiRYuyzjrr3O5r3XXXzXrrrZfFixdn8eLFSZIlS5YM9w8yQ6xoBwAAAABgwK9+9avcdtttnev8+te/7qGb0WdFOwAAAAAAAy6//PJe6lx77bW91Bl1VrQDAAAAADBg3333zb777tupxtjYWB73uMf11NFos6IdAAAAAAA6ELQDAAAAAEAHgnYAAAAAAOhA0A4AAAAAAB0I2gEAAAAAoINFw24AAAAAAIDRcvPNN+cTn/hEaq2ptSbJ6ttJk/umb5/cNz4+nttuu21G+h02QTsAAAAAAAOOP/74nHHGGZ3rnHHGGTnssMN66Gi0CdoBAAAAABhw8MEHZ9NNN139uJQysL+UsnrbHe076aSTsscee6z9ZkeAoB0AAAAAgAEbbrhhXvSiF3Wqccopp2ThwoU9dTTaXAwVAAAAAAA6ELQDAAAAAEAHgnYAAAAAAOhA0A4AAAAAAB24GCoAAAAAAANuuummnHLKKZmYmFi9rdZ6l24n74+Pj+fWW2+dqZaHStAOAAAAAMCAd73rXTnjjDM61znzzDNzxBFH9NDRaBO0AwAAAAAw4NBDD8197nOf1Y9LKWu8naqUMrD9+OOPz5577rmWOx0NgnYAAAAAAAasv/76ee5zn9upxvve974sWDA/LhM6P/6UAAAAAACwlgjaAQAAAACgA0E7AAAAAAB0YEY7AAAAAAADvvvd7+aII47oXOfqq6/uoZvRZ0U7AAAAAAADvvvd7/ZS58orr+ylzqizoh0AAAAAgAH7779/9t9//041xsbGsv322/fU0Wizoh0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHbgYKgAAAAAAAz75yU/m7W9/e+c65513XpYuXdpDR6PNinYAAAAAAAbc5z736aXOn/3Zn/VSZ9RZ0Q4AAAAAwIDHPvaxWb58eacaY2Nj2XLLLftpaMRZ0Q4AAAAAAB0I2gEAAAAAoAOjYwAAAAAAGHDaaafl+OOP71znnHPOmRcXQxW0AwAAMOctWbIkK1asGHYbzKCVK1cOuwWAWe0HP/hBL3V+/vOf91Jn1AnaAQAAmPNWrlyZI488cthtMEOOPfbYYbcAMOs96UlPype+9KXOdR784Af30M3oM6MdAAAAAIABF1xwQS91Lr300l7qjDor2gEAAAAAGLDPPvvkoosuysTERGqtA/smH9daB/ZNv3/FFVfk8Y9//Mw0PGSCdgAAAAAABvz5n/95TjjhhE41xsbGsskmm/TU0WgzOgYAAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHQjaAQAAAACgA0E7AAAAAAB0IGgHAAAAAIAOFg27AQAAAAAARss3vvGNHH300Z3rXH755T10M/qsaAcAAAAAYMBVV13VS53rr7++lzqjzop2AAAAAAAGPOtZz8qznvWsTjXGxsay3Xbb9dTRaLOiHQAAAAAAOhC0AwAAAABAB0bHAAAAAAAwoNaaX/7yl6vvr+l28v6atifJqlWrZqLVkSBoBwAAAABgwHvf+9585CMf6Vzn85//fJYuXdpDR6NN0A4AAAAAwICnP/3p+fznP5+JiYkkSSlljbd3tu/aa6/NYx/72BnreZgE7QAAAAAADNhyyy1z2mmndaoxNjaWu9/97v00NOJcDBUAAAAAADoQtAMAAAAAQAdGxwAAAAAAMKDWmp/+9KeptabWert9k7dT902/f+utt85MsyNA0A4AAAAAwIB3v/vd+djHPta5zuc+97ksXbq0h45Gm9ExAAAAAAAMuP7663upc8stt/RSZ9QJ2gEAAAAAGLDvvvv2UucJT3hCL3VGndExAAAAAAAM2GabbbJ8+fJONcbGxrLZZpv11NFoE7QDAAAAADDg17/+dU4++eSsWrVq9bY1Xfh0+oVSp24bHx+fNxdEFbQDAAAAADDgPe95T5YtW9a5zplnnpkjjjiih45Gm6AdAAAAAIABhx56aLbccsvVj0spA7dTTd03df/b3va27Lnnnmu30REhaAcAAAAAYMDixYuzzz77dKpxwgknZMGCBT11NNrmx58SAAAAAADWEkE7AAAAAAB0IGgHAAAAAIAOBO0AAAAAANCBoB0AAAAAADoQtAMAAAAAQAeCdgAAAAAA6EDQDgAAAAAAHSwadgMAAACwti1ZsiQrVqwYdhvMoJUrVw67BYBZ7dRTT80JJ5zQuc65556bpUuX9tDRaBO0AwAAMOetXLkyRx555LDbYIYce+yxw24BYNY79dRTe6lz5ZVX9lJn1AnaAQAAAAAYcPLJJ68O20spA/tKKau/pm+fuv8973lP9t1335lpeMgE7QAAAAAADNhggw3yghe8oFONk08+OQsXLuynoRHnYqgAAAAAANCBFe0AAAAAAKxRrXX115oeT0xM3OG+ydv5QNAOAAAAAMCAt7zlLTnrrLM61/n0pz+dpUuX9tDRaDM6BgAAAACAAT/60Y96qXPzzTf3UmfUCdoBAAAAABiwzTbb9FJn8eLFvdQZdUbHAAAAAAAw4NBDD83mm28+MGd9+uz1O7qdvH/KKadk9913n6mWh0rQDgAAAADAgA022CAHHnhgpxqnn356Fi2aHxG00TEAAAAAANCBoB0AAAAAADqYH+v2AQAAAAC4y84777y84hWv6Fznpz/9aQ/djD4r2gEAAAAAGHD11Vf3UufnP/95L3VGnRXtAAAAAAAMeOYzn5lnPvOZnWrsuuuu2XbbbXvqaLRZ0Q4AAAAAQO9KKcNuYcYI2gEAAAAAoAOjYwAAAAAAGHDFFVfk5S9/eVatWpVaa5Ksvp1Ua139NV2tNePj4/nFL34xI/0Om6AdAAAAAIABn/vc53q5kOm3v/3tHroZfYJ2AAAAAAAG/P3f/33222+/1XPWp99ONblt+sr25zznOXnqU5+6ljsdDYJ2AAAAAAAGlFJyj3vco3ON+ULQDgAAAADAgFprbrrppjXOYJ86s33qvun3JyYmZqbZESBoBwAAAABgwIknnphTTjmlc50vfOELWbp0aQ8djTZBOwAAAAAAA572tKdl2bJlq1epTx8DU0q5w7ntk/t+/vOf5zGPeczMNDxkgnYAAAAAAAbc7373y2te85o1jo5JBsfHTN826VWvelXufve7r9U+R4WgHQAAAACAAccff3w++clPdq5zxhln5OCDD+6ho9EmaAcAAAAAYMBzn/vcXHXVVbfbvqZxMdPvTz7+2te+lqc+9alrudPRIGgHAAAAAGDAJZdckvPPP79znf/7v//roZvRt2DYDQAAAAAAMFquv/76XurcdNNNvdQZdVa0AwAAAAAwYM8998yee+7ZqcbY2Fge+chH9tTRaLOiHQAAAAAAOhC0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdCNoBAAAAAKADQTsAAAAAAHQgaAcAAAAAgA4E7QAAAAAA0IGgHQAAAAAAOhC0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdLBp2AwAAAAAAjJZLLrkkhx9+eFatWrV6W631Lt1Odd11163tVkeCoB0AAAAAgAErVqzI+Ph45zoXXnhhD92MPkE7AAAAAAADXvSiF+WAAw5IKWWN+2utd7iSffLxs571rDzlKU9Zu42OCEE7AAAAAAADSilZb731OteYLwTtAAAAAAAM+N73vpdDDz20c52f/vSnPXQz+hYMuwEAAAAAAEbLN7/5zV7qXHzxxb3UGXVWtAMAAAAAMOCAAw7IAQcc0KnGrrvumh133LGnjkabFe0AAAAAAPTOjHYAAAAAAOat5cuX5/Wvf33nOt/73vd66Gb0WdEOAAAAAMCA3/3ud73UmZiY6KXOqLOiHQAAAACAAU9/+tPz9Kc/vVONsbGxPPShD+2po9FmRTsAAAAAAHQgaAcAAAAAgA4E7QAAAAAA0IGgHQAAAAAAOhC0AwAAAABAB4J2AAAAAADoQNAOAAAAAAAdLBp2AwAAAAAAjJZaa66++urUWlNrvd2+ydup+6bfHx8fn5lmR4CgHQAAAACAAe9617ty6qmndq5z1llnZenSpT10NNoE7QAAAAAADNhjjz3yta99bfUq9VLKwP5Syhq3Td132WWXZfvtt5+ZhodM0A4AAAAAwIC/+Iu/yMknn9ypxtjYWDbeeOOeOhptgnYAAAAAAAZccsklOeyww7Jq1aqBmezT3dGM9knXXXfd2mtyhAjaAQAAAAAY8JWvfCW33HJL5zoXXnhhD92MPkE7AAAAAAADXvSiF+WAAw643Rz2Na1gv6MV78985jPzlKc8ZS13OhoE7QAAAAAA3M66667b6fXTQ/q5bMGwGwAAAAAAgNlM0A4AAAAAAB0YHQMAAAAAwIBVq1blhz/8YZI1z2CfPo99Tc8bHx9f222ODEE7AAAAAAADTjjhhHziE5/oXOezn/1sli5d2kNHo83oGAAAAAAABmyyySa91Nlss816qTPqBO0AAAAAAAw46aSTeqlz2WWX9VJn1AnaAQAAAAAY8OpXv7qXOo94xCN6qTPqBO0AAAAAAAy45ZZbeqlz66239lJn1AnaAQAAAAAYcO211/ZS5/rrr++lzqgTtAMAAAAAMGCHHXa40/2llCxYsGDga+HChVm4cGEWLVqUddZZJ0ny8Ic/fCbaHbpFw24AAAAAAIDRsvXWW2f58uWdaoyNjeVe97pXTx2NNivaAQAAAACgAyvaAQAAAAAY8L3vfS+HHnpo5zo/+clPeuhm9FnRDgAAAADAgG9+85u91Lnkkkt6qTPqrGgHAAAAAGDAAQcckAMOOKBTjbGxsey44449dTTarGgHAAAAAIAOrGgHAAAAAGDAV7/61fzzP/9z5zo//OEPe+hm9FnRDgAAAADAgBtuuKGXOuPj473UGXWCdgAAAAAABlx22WW91Lnqqqt6qTPqjI4BAAAAAGDA/vvvn2uuuSa11iRJKWXgdqqp+6buP+ecc/LUpz51BrodPkE7AAAAAAADNtpoozzsYQ9LrXV12D55O9XUbdOfd84552TdddedgW6HT9AOAAAAAMCAd77znVm2bFnnOsuWLcuhhx7aQ0ejTdAOAAAAAMCAF77whTn77LOzatWqO33e9FXuU0fH/OY3v8luu+22VvobNYJ2AAAAAAAGvP/978+vfvWrznU++9nP5vDDD++ho9EmaAcAAAAAYMBBBx2UddZZ53YXQ5009cKnd7Tv1FNPtaIdAAAAAID5aeONN84RRxzxR71m8sKpk+H8pz71qayzzjpro72RI2gHAAAAAGDAhz/84Zx44omd65x99tlZunRpDx2NtgXDbgAAAAAAgNHyiEc8opc6W2+9dS91Rp0V7QAAAAAADHjoQx+a5cuXd6oxNjaWLbbYoqeORpugHQAAgDlvyZIlWbFixbDbYAatXLly2C0AzGpvectbctZZZ3Wu85nPfGZejI4RtAMAADDnrVy5MkceeeSw22CGHHvsscNuAWDW+/GPf9xLnd/+9re91Bl1gnYAAAAAAAa84x3vyE033ZRSysD2ycellNRak2T17aTJx/vtt1/23nvvtd/sCBC0AwAAAAAwoJSSu93tbp1rzBcLht0AAAAAAABzS631divd5zIr2gEAAAAAGPCRj3wk73nPezrX+epXvzovLoZqRTsAAAAAAAMe+tCH9lLngQ98YC91Rp0V7QAAAAAADHjYwx6W5cuXd6oxNjaWLbbYoqeORpugHQAAAACAAbfccks+97nPZWJi4naz1icfT5/DPv3++Ph4JiYmZqbhIRO0AwAAAAAw4Pjjj89nPvOZznU+85nP5JBDDumho9EmaAcAAAAAYMCLX/ziLFiwYPUq9VLKGm+nKqWs/kqS0047LbvtttsMdTxcgnYAAAAAAAZsvPHGOfLIIzvVWLZsWdZZZ51+GhpxC4bdAAAAAAAAzGZ3KWgvpexbSjmulPLVUsqvSim1lPJfd/DcLdv9d/T10Tv57zy/lHJeKeWmUsqNpZQVpZQ97uT5C0spR5VSLiil3FxK+UUp5bOllO3v5DXrl1JeV0r5YSllvJTys1LKqaWUB9+V/xcAAAAAAF2VUu5WSjm2lHJlm22eU0rZbth9zWWllCeVUj5dSrm6zapfMG1/KaUcU0q5pj0mK0op296V2nd1dMyrkzwiyU1J/i/JNnfhNd9J8t9r2H7hmp5cSvn3JC9r6783ybpJ/ibJZ0oph9da3zHt+SXJR5Psm+SHSd6R5B5JnpPk7FLKPrXWT017zXpJvpBkhyTnJ/nPJH+RZL8ku5dSdqq1fv0u/NkAAAAAALo4McnDkzw/TSb6t0m+WEp5yORc9GG69NJLc9hhh+W2225LrXXg649x4403rqUO/yQbpcmnT26/pjs6TUb9gjSZ8z8n+UIp5UG11l/fWeG7GrQfleZgX5LkyUmW34XXrKy1HnNXircr0F+W5NIk29Vaf9lu/7ck30zy76WUZbXWK6a87G/ShOznJHlqrXW8fc27kvxPkveWUr487X/AS9OE7KcleU6tdaJ9zcfSfChwUinlYZPbAQAAAAD6VkpZP8k+Sfapta5oNx9TStkzycFDa6y1atWqvPjFL+6l1llnnZUnPelJeexjH5uFCxf2UvNPVWv9bJLPJkkp5QNT97ULu49M8q+11k+0256f5GdJ9k/y7jurfZdGx9Ral9daL65r76OUpe3tv0yG7O1/94okxydZL8kLp71m8i/cqydD9vY130jysSSbpQnik6z+HzX53zl6apjernz/apKHpPkgAQAAAABgbVmUZGGS8Wnbb07yxJlv5/dWrVqVo48+Ouutt15vNd/whjfk6KOPzqpVq3qruRZslWSLJJ+f3FBrvTnJ2UnucFT5pLV5MdR7l1JeUkp5VXv78Dt57k7t7Vlr2HfmtOeklLI4zR/ut2kC8j/4miQPSHK/JD+qtV5+F18DAAAAANCrdgrH15K8upRyn/ZalH+b5AlJ/nyYvZ133nn5/ve/n1tuuaW3mjfffHO+973v5bzzzuut5lqwRXt77bTt107Zd4fKH7tIvZTy12lGx3y41vq3a9i/ZZI1BdlJsiLJ82utP57y/A3TzH6/qdZ6tzXUu1eS65L8rNa6ebtt2zSzdC6stT5sDa95TJJvJDmv1vq4dtvuSZYlWVZr3XMNr9k3yceTnFprfc4d9A8AAAAA0Fkp5QFJTkrypCSrknwryY+SPLrW+uBh9fWUpzzlNUmOyeAi7Ykkr12+fPkbZ6rG2lZKuSnJYbXWD7SPt0/yv0n+clp+fVKS+9Rad7mzend1Rvsf47dJ3pBm5vll7baHp/kf+5QkXyqlLKm1/qbdt0l7e0dT8Se3333Ktpl6DQAAAABA72qtlyZ5crsQeeNa60/aa0le9gdeulYtX778DWny3aHWGIKftrebJ/nxlO2bT9l3h3ofHVNr/Vmt9Z9rrd+qtd7Qfp2d5OlJvp7kr5L0M0kfAAAAAGAWq7X+pg3ZN02yS5JPDbuneeryNIH6zpMb2hHmOyY55w+9eG3OaB9Qa70tyYntwydN2TW5knyTrNnk9huG8BoAAAAAgN6VUnYppYyVUrYqpeycZlz3D5K8f8itzVmllI1KKUtKKUvSZOP3ax/frzYz1o9N8opSyrNKKQ9N8oE0Y89P+UO1Zyxob13X3m44uaEdIXN1ko1KKWsa9P/A9vZHU7ZdmmZu0f1LKWsaf7Om1/ywvd36Dnpb02sAAAAAANaGTZK8I024fnKS/0myS631d0Ptam57TJJvt1/rJ3lde//17f63JnlbkuOTnJ/mwrRPby9ee6dmOmh/fHs7fc7Ql9vbXdfwmrFpz0mtdTzNcv0N0izd/4OvSRPO/zjJ1qWUre7iawAAAAAAeldrPbXW+oBa63q11j+vtR5Wa72j60vSg1rrilprWcPXC9r9tdZ6THs8Ftdan1xrvfCu1O49aC+lPKqUcru6pZSnJjmqffhf03a/q739p3YW0eRrtkxyaJJbcvtTJk5ob9/YzsqZfM12SZ6TZvX8Jya3t0v/J/87b53aYynlGWkC++8l+cof/lMCAAAAAECjNPnzH3hSKXsn2bt9uEWaofyXJflqu+36WuvL2+euSDOG5Zwk/9fuf3iSndr7r6m1vnEN/43/SPLS9jWnJVk3TWB+zySH11rfMe35JcmpSfZNc3rFZ9rnPifJ4iT71Fo/Ne0166VZsb59mqX/X0pyvyT7Jbk1yU611q//wf8hAAAAAADQuqtB+zFJXnsnT7my1rpl+9wXJXlmkocmuVeSdZJcm+RrSd5Ra/3qHRUppbwgzQr2hySZSPKtJP9Wa112B89flOTwJH+X5K+SjLf/nTfWWtd4JdhSygZJXpnkuWlC9l8lWZHktbXW793JnxEAAAAAAG7nLgXtAAAAAADAms30xVABAAAAAGBOEbQDAAAAAEAHgnYAAAAAAOhA0A4AAAAAAB0I2gEAAAAAoANBOwAAAAAAdCBoBwAAAACADgTtAAAAAADQgaAdAAAAAAA6+P8BGtVt7o1drMEAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1800x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# The msno.matrix nullity matrix is a data-dense display which lets us quickly visually pick out patterns in data completion.\n",
    "# Check known.csv dataset.\n",
    "msno.matrix(known_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "f6           146\n",
       "f2             1\n",
       "longitude      0\n",
       "latitude       0\n",
       "f1             0\n",
       "f3             0\n",
       "f4             0\n",
       "f5             0\n",
       "t1             0\n",
       "id             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Missing values\n",
    "known_dataset.isnull().sum().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>count</th>\n",
       "      <th>percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>f2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f6</th>\n",
       "      <td>146</td>\n",
       "      <td>0.97</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    count  percentage\n",
       "f2      1        0.01\n",
       "f6    146        0.97"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Summation of the missing values and calculation of the missing values as a percentage in known dataset.\n",
    "missing_values = pd.DataFrame(known_dataset.isnull().sum(),columns=['count'])\n",
    "missing_values['percentage'] = (missing_values['count']/known_dataset.shape[0]*100).round(2)\n",
    "missing_values.query('count != 0')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# The msno.matrix nullity matrix is a data-dense display which lets us quickly visually pick out patterns in data completion.\n",
    "# Check unknown.csv dataset.\n",
    "msno.matrix(unknown_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "f6           9\n",
       "longitude    0\n",
       "latitude     0\n",
       "f1           0\n",
       "f2           0\n",
       "f3           0\n",
       "f4           0\n",
       "f5           0\n",
       "t1           0\n",
       "id           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Missing values\n",
    "unknown_dataset.isnull().sum().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>count</th>\n",
       "      <th>percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>f6</th>\n",
       "      <td>9</td>\n",
       "      <td>1.17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    count  percentage\n",
       "f6      9        1.17"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Summation of the missing values and calculation of the missing values as a percentage in unknown dataset.\n",
    "missing_values = pd.DataFrame(unknown_dataset.isnull().sum(),columns=['count'])\n",
    "missing_values['percentage'] = (missing_values['count']/unknown_dataset.shape[0]*100).round(2)\n",
    "missing_values.query('count != 0')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`So we decide to impute missing values with Mean/Median, because columns in the dataset which are having numeric continous values can be replaced with the mean, median, or mode of remaining values in the column. This method can prevent the loss of data. Replacing two approximations (mean, median) is a statistical approach to handle the missing values.`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "known_dataset['f6'] = known_dataset['f6'].replace(np.NaN, known_dataset['f6'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "known_dataset['f2'] = known_dataset['f2'].replace(np.NaN, known_dataset['f2'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "longitude    0\n",
       "latitude     0\n",
       "f1           0\n",
       "f2           0\n",
       "f3           0\n",
       "f4           0\n",
       "f5           0\n",
       "f6           0\n",
       "t1           0\n",
       "id           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check again for missing values in known.csv dataset.\n",
    "known_dataset.isnull().sum().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "unknown_dataset['f6'] = unknown_dataset['f6'].replace(np.NaN, unknown_dataset['f6'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "longitude    0\n",
       "latitude     0\n",
       "f1           0\n",
       "f2           0\n",
       "f3           0\n",
       "f4           0\n",
       "f5           0\n",
       "f6           0\n",
       "t1           0\n",
       "id           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check again for missing values in unknown.csv dataset.\n",
    "unknown_dataset.isnull().sum().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Export Cleaned Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save cleaned known dataset to a new .csv file\n",
    "path = 'data/cleaned_known.csv'\n",
    "known_dataset.to_csv(path, \n",
    "index=False) # Avoid creating an index in a saved .csv file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save cleaned unknown dataset to a new .csv file\n",
    "path = 'data/cleaned_unknown.csv'\n",
    "unknown_dataset.to_csv(path, \n",
    "index=False) # Avoid creating an index in a saved .csv file."
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "79aec032834ca8c7adc23337e31b2f95bcd74aaeb73f4918421d2e03d3d0781b"
  },
  "kernelspec": {
   "display_name": "Python 3.8.5 64-bit ('.venv': venv)",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}