Capstone.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Loading Hive Tables and Data Preparation for Analysis**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <p><b>SparkSession - hive</b></p>\n",
       "                \n",
       "        <div>\n",
       "            <p><b>SparkContext</b></p>\n",
       "\n",
       "            <p><a href=\"http://ip-10-1-1-204.ap-south-1.compute.internal:43002\">Spark UI</a></p>\n",
       "\n",
       "            <dl>\n",
       "              <dt>Version</dt>\n",
       "                <dd><code>v2.4.0</code></dd>\n",
       "              <dt>Master</dt>\n",
       "                <dd><code>local[*]</code></dd>\n",
       "              <dt>AppName</dt>\n",
       "                <dd><code>Subham_Capstone</code></dd>\n",
       "            </dl>\n",
       "        </div>\n",
       "        \n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "<pyspark.sql.session.SparkSession at 0x7fa2f087c1d0>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pyspark.sql import SparkSession\n",
    "\n",
    "spark = SparkSession.builder.appName(\"Subham_Capstone\")\\\n",
    "    .config(\"hive.metastore.uris\",\"thrift://ip-10-1-2-24.ap-south-1.compute.internal:9083\")\\\n",
    "    .config(\"spark.sql.catalogImplementation=hive\")\\\n",
    "    .config(\"spark.sql.warehouse.dir\", \"hdfs://nameservice1/user/anabig114212/hive/Capstone\")\\\n",
    "    .config(\"spark.serializer\", \"org.apache.spark.serializer.KryoSerializer\")\\\n",
    "    .enableHiveSupport().getOrCreate()\n",
    "spark"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing Libraraies for Analysis\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime as dt\n",
    "import seaborn as sns\n",
    "import re\n",
    "from itertools import combinations as cm\n",
    "\n",
    "#import matplotlib.ticker as ticker\n",
    "import matplotlib.dates as mdates\n",
    "\n",
    "# set the graphs to show in the jupyter notebook\n",
    "%matplotlib inline\n",
    "\n",
    "# set seabor graphs to a better style\n",
    "sns.set(style='ticks')\n",
    "\n",
    "from pyspark.sql.functions import *\n",
    "import pyspark.sql.functions as F\n",
    "\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating Local Views and Spark Dataframes to call these objects from memory.\n",
    "\n",
    "departments = spark.table('anabig114212_cap.departments')\n",
    "departments.createOrReplaceTempView('departments')\n",
    "\n",
    "titles = spark.table('anabig114212_cap.titles')\n",
    "titles.createOrReplaceTempView('titles')\n",
    "\n",
    "employees = spark.table('anabig114212_cap.employees')\n",
    "employees = employees.withColumn('birth_date', to_date(col('birth_date'),'yyyy-MM-dd'))\\\n",
    "                .withColumn('hire_date', to_date(col('hire_date'),'yyyy-MM-dd'))\\\n",
    "                .withColumn('last_date', to_date(col('last_date'),'yyyy-MM-dd')) #Converting to Proper date formats\n",
    "employees.createOrReplaceTempView('employees')\n",
    "\n",
    "dept_emp_raw = spark.table('anabig114212_cap.dept_emp')\n",
    "dept_emp = spark.table('anabig114212_cap.dept_emp1')\n",
    "dept_emp.createOrReplaceTempView('dept_emp')\n",
    "\n",
    "dept_manager = spark.table('anabig114212_cap.dept_manager')\n",
    "dept_manager.createOrReplaceTempView('dept_manager')\n",
    "\n",
    "salaries = spark.table('anabig114212_cap.salaries')\n",
    "salaries.createOrReplaceTempView('salaries')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Validating Data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------+------------------+\n",
      "|dept_no|         dept_name|\n",
      "+-------+------------------+\n",
      "|   d001|         Marketing|\n",
      "|   d002|           Finance|\n",
      "|   d003|   Human Resources|\n",
      "|   d004|        Production|\n",
      "|   d005|       development|\n",
      "|   d006|Quality Management|\n",
      "|   d007|             Sales|\n",
      "|   d008|          Research|\n",
      "|   d009|  Customer Service|\n",
      "+-------+------------------+\n",
      "\n",
      "root\n",
      " |-- dept_no: string (nullable = true)\n",
      " |-- dept_name: string (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "departments.show()\n",
    "departments.printSchema()\n",
    "departments.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------+------------------+\n",
      "|title_id|             title|\n",
      "+--------+------------------+\n",
      "|   e0001|Assistant Engineer|\n",
      "|   e0002|          Engineer|\n",
      "|   e0003|   Senior Engineer|\n",
      "|   e0004|  Technique Leader|\n",
      "|   m0001|           Manager|\n",
      "|   s0001|             Staff|\n",
      "|   s0002|      Senior Staff|\n",
      "+--------+------------------+\n",
      "\n",
      "root\n",
      " |-- title_id: string (nullable = true)\n",
      " |-- title: string (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "titles.show()\n",
    "titles.printSchema()\n",
    "titles.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- emp_no: integer (nullable = true)\n",
      " |-- emp_title_id: string (nullable = true)\n",
      " |-- birth_date: date (nullable = true)\n",
      " |-- first_name: string (nullable = true)\n",
      " |-- last_name: string (nullable = true)\n",
      " |-- sex: string (nullable = true)\n",
      " |-- hire_date: date (nullable = true)\n",
      " |-- no_of_projects: integer (nullable = true)\n",
      " |-- last_performance_rating: string (nullable = true)\n",
      " |-- left2: integer (nullable = true)\n",
      " |-- last_date: date (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "300024"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "employees.printSchema()\n",
    "employees.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>emp_no</th>\n",
       "      <th>emp_title_id</th>\n",
       "      <th>birth_date</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>sex</th>\n",
       "      <th>hire_date</th>\n",
       "      <th>no_of_projects</th>\n",
       "      <th>last_performance_rating</th>\n",
       "      <th>left2</th>\n",
       "      <th>last_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10001</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-09-02</td>\n",
       "      <td>Georgi</td>\n",
       "      <td>Facello</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-06-26</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "      <td>1994-07-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10002</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1964-06-02</td>\n",
       "      <td>Bezalel</td>\n",
       "      <td>Simmel</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-11-21</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10003</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1959-12-03</td>\n",
       "      <td>Parto</td>\n",
       "      <td>Bamford</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-08-28</td>\n",
       "      <td>1</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10004</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1954-05-01</td>\n",
       "      <td>Chirstian</td>\n",
       "      <td>Koblick</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-12-01</td>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10005</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1955-01-21</td>\n",
       "      <td>Kyoichi</td>\n",
       "      <td>Maliniak</td>\n",
       "      <td>M</td>\n",
       "      <td>1989-09-12</td>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>10006</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-04-20</td>\n",
       "      <td>Anneke</td>\n",
       "      <td>Preusig</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-06-02</td>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10007</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1957-05-23</td>\n",
       "      <td>Tzvetan</td>\n",
       "      <td>Zielinski</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-02-10</td>\n",
       "      <td>6</td>\n",
       "      <td>B</td>\n",
       "      <td>1</td>\n",
       "      <td>2002-09-18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10008</td>\n",
       "      <td>e0001</td>\n",
       "      <td>1958-02-19</td>\n",
       "      <td>Saniya</td>\n",
       "      <td>Kalloufi</td>\n",
       "      <td>M</td>\n",
       "      <td>1994-09-15</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>10009</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1952-04-19</td>\n",
       "      <td>Sumant</td>\n",
       "      <td>Peac</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-02-18</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10010</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1963-06-01</td>\n",
       "      <td>Duangkaew</td>\n",
       "      <td>Piveteau</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-08-24</td>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10011</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1953-11-07</td>\n",
       "      <td>Mary</td>\n",
       "      <td>Sluis</td>\n",
       "      <td>F</td>\n",
       "      <td>1990-01-22</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>10012</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1960-10-04</td>\n",
       "      <td>Patricio</td>\n",
       "      <td>Bridgland</td>\n",
       "      <td>M</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>3</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>10013</td>\n",
       "      <td>s0002</td>\n",
       "      <td>1963-06-07</td>\n",
       "      <td>Eberhardt</td>\n",
       "      <td>Terkki</td>\n",
       "      <td>M</td>\n",
       "      <td>1985-10-20</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>10014</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1956-02-12</td>\n",
       "      <td>Berni</td>\n",
       "      <td>Genin</td>\n",
       "      <td>M</td>\n",
       "      <td>1987-03-11</td>\n",
       "      <td>9</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>10015</td>\n",
       "      <td>s0002</td>\n",
       "      <td>1959-08-19</td>\n",
       "      <td>Guoxiang</td>\n",
       "      <td>Nooteboom</td>\n",
       "      <td>M</td>\n",
       "      <td>1987-07-02</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10016</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1961-05-02</td>\n",
       "      <td>Kazuhito</td>\n",
       "      <td>Cappelletti</td>\n",
       "      <td>M</td>\n",
       "      <td>1995-01-27</td>\n",
       "      <td>5</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>10017</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1958-07-06</td>\n",
       "      <td>Cristinel</td>\n",
       "      <td>Bouloucos</td>\n",
       "      <td>F</td>\n",
       "      <td>1993-08-03</td>\n",
       "      <td>10</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10018</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1954-06-19</td>\n",
       "      <td>Kazuhide</td>\n",
       "      <td>Peha</td>\n",
       "      <td>F</td>\n",
       "      <td>1987-04-03</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10019</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1953-01-23</td>\n",
       "      <td>Lillian</td>\n",
       "      <td>Haddadi</td>\n",
       "      <td>M</td>\n",
       "      <td>1999-04-30</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10020</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1952-12-24</td>\n",
       "      <td>Mayuko</td>\n",
       "      <td>Warwick</td>\n",
       "      <td>M</td>\n",
       "      <td>1991-01-26</td>\n",
       "      <td>7</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    emp_no emp_title_id  birth_date first_name    last_name sex   hire_date  \\\n",
       "0    10001        e0003  1953-09-02     Georgi      Facello   M  1986-06-26   \n",
       "1    10002        s0001  1964-06-02    Bezalel       Simmel   F  1985-11-21   \n",
       "2    10003        e0003  1959-12-03      Parto      Bamford   M  1986-08-28   \n",
       "3    10004        e0003  1954-05-01  Chirstian      Koblick   M  1986-12-01   \n",
       "4    10005        s0001  1955-01-21    Kyoichi     Maliniak   M  1989-09-12   \n",
       "5    10006        e0003  1953-04-20     Anneke      Preusig   F  1989-06-02   \n",
       "6    10007        s0001  1957-05-23    Tzvetan    Zielinski   F  1989-02-10   \n",
       "7    10008        e0001  1958-02-19     Saniya     Kalloufi   M  1994-09-15   \n",
       "8    10009        e0003  1952-04-19     Sumant         Peac   F  1985-02-18   \n",
       "9    10010        e0002  1963-06-01  Duangkaew     Piveteau   F  1989-08-24   \n",
       "10   10011        s0001  1953-11-07       Mary        Sluis   F  1990-01-22   \n",
       "11   10012        e0003  1960-10-04   Patricio    Bridgland   M  1992-12-18   \n",
       "12   10013        s0002  1963-06-07  Eberhardt       Terkki   M  1985-10-20   \n",
       "13   10014        e0002  1956-02-12      Berni        Genin   M  1987-03-11   \n",
       "14   10015        s0002  1959-08-19   Guoxiang    Nooteboom   M  1987-07-02   \n",
       "15   10016        s0001  1961-05-02   Kazuhito  Cappelletti   M  1995-01-27   \n",
       "16   10017        s0001  1958-07-06  Cristinel    Bouloucos   F  1993-08-03   \n",
       "17   10018        e0003  1954-06-19   Kazuhide         Peha   F  1987-04-03   \n",
       "18   10019        s0001  1953-01-23    Lillian      Haddadi   M  1999-04-30   \n",
       "19   10020        e0002  1952-12-24     Mayuko      Warwick   M  1991-01-26   \n",
       "\n",
       "    no_of_projects last_performance_rating  left2   last_date  \n",
       "0                9                       C      1  1994-07-30  \n",
       "1                8                       B      0        None  \n",
       "2                1                       C      0        None  \n",
       "3                5                       A      0        None  \n",
       "4                6                       A      0        None  \n",
       "5               10                       B      0        None  \n",
       "6                6                       B      1  2002-09-18  \n",
       "7                9                       C      0        None  \n",
       "8                8                       B      0        None  \n",
       "9                4                       A      0        None  \n",
       "10               6                       C      0        None  \n",
       "11               3                       A      0        None  \n",
       "12               6                       C      0        None  \n",
       "13               9                       B      0        None  \n",
       "14               4                       B      0        None  \n",
       "15               5                       C      0        None  \n",
       "16              10                       A      0        None  \n",
       "17               4                       B      0        None  \n",
       "18               4                       B      0        None  \n",
       "19               7                       B      0        None  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "employees.toPandas().head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-------+\n",
      "|emp_no|dept_no|\n",
      "+------+-------+\n",
      "| 10001|   d005|\n",
      "| 10002|   d007|\n",
      "| 10003|   d004|\n",
      "| 10004|   d004|\n",
      "| 10005|   d003|\n",
      "| 10006|   d005|\n",
      "| 10007|   d008|\n",
      "| 10008|   d005|\n",
      "| 10009|   d006|\n",
      "| 10010|   d004|\n",
      "| 10010|   d006|\n",
      "| 10011|   d009|\n",
      "| 10012|   d005|\n",
      "| 10013|   d003|\n",
      "| 10014|   d005|\n",
      "| 10015|   d008|\n",
      "| 10016|   d007|\n",
      "| 10017|   d001|\n",
      "| 10018|   d004|\n",
      "| 10018|   d005|\n",
      "+------+-------+\n",
      "only showing top 20 rows\n",
      "\n",
      "root\n",
      " |-- emp_no: integer (nullable = true)\n",
      " |-- dept_no: string (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "331603"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dept_emp_raw.show()\n",
    "dept_emp_raw.printSchema()\n",
    "dept_emp_raw.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### dept_emp table was created by removing duplicate emp_no, to keep only the last the department for each employee\n",
    "[Created in Hive]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-------+\n",
      "|emp_no|dept_no|\n",
      "+------+-------+\n",
      "| 10001|   d005|\n",
      "| 10002|   d007|\n",
      "| 10003|   d004|\n",
      "| 10004|   d004|\n",
      "| 10005|   d003|\n",
      "| 10006|   d005|\n",
      "| 10007|   d008|\n",
      "| 10008|   d005|\n",
      "| 10009|   d006|\n",
      "| 10010|   d006|\n",
      "| 10011|   d009|\n",
      "| 10012|   d005|\n",
      "| 10013|   d003|\n",
      "| 10014|   d005|\n",
      "| 10015|   d008|\n",
      "| 10016|   d007|\n",
      "| 10017|   d001|\n",
      "| 10018|   d005|\n",
      "| 10019|   d008|\n",
      "| 10020|   d004|\n",
      "+------+-------+\n",
      "only showing top 20 rows\n",
      "\n",
      "root\n",
      " |-- emp_no: integer (nullable = true)\n",
      " |-- dept_no: string (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "300024"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dept_emp.show()\n",
    "dept_emp.printSchema()\n",
    "dept_emp.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------+------+\n",
      "|dept_no|emp_no|\n",
      "+-------+------+\n",
      "|   d001|110022|\n",
      "|   d001|110039|\n",
      "|   d002|110085|\n",
      "|   d002|110114|\n",
      "|   d003|110183|\n",
      "|   d003|110228|\n",
      "|   d004|110303|\n",
      "|   d004|110344|\n",
      "|   d004|110386|\n",
      "|   d004|110420|\n",
      "|   d005|110511|\n",
      "|   d005|110567|\n",
      "|   d006|110725|\n",
      "|   d006|110765|\n",
      "|   d006|110800|\n",
      "|   d006|110854|\n",
      "|   d007|111035|\n",
      "|   d007|111133|\n",
      "|   d008|111400|\n",
      "|   d008|111534|\n",
      "+-------+------+\n",
      "only showing top 20 rows\n",
      "\n",
      "root\n",
      " |-- dept_no: string (nullable = true)\n",
      " |-- emp_no: integer (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "24"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dept_manager.show()\n",
    "dept_manager.printSchema()\n",
    "dept_manager.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+------+\n",
      "|emp_no|salary|\n",
      "+------+------+\n",
      "| 10001| 60117|\n",
      "| 10002| 65828|\n",
      "| 10003| 40006|\n",
      "| 10004| 40054|\n",
      "| 10005| 78228|\n",
      "| 10006| 40000|\n",
      "| 10007| 56724|\n",
      "| 10008| 46671|\n",
      "| 10009| 60929|\n",
      "| 10010| 72488|\n",
      "| 10011| 42365|\n",
      "| 10012| 40000|\n",
      "| 10013| 40000|\n",
      "| 10014| 46168|\n",
      "| 10015| 40000|\n",
      "| 10016| 70889|\n",
      "| 10017| 71380|\n",
      "| 10018| 55881|\n",
      "| 10019| 44276|\n",
      "| 10020| 40000|\n",
      "+------+------+\n",
      "only showing top 20 rows\n",
      "\n",
      "root\n",
      " |-- emp_no: integer (nullable = true)\n",
      " |-- salary: integer (nullable = true)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "300024"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "salaries.show()\n",
    "salaries.printSchema()\n",
    "salaries.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preparing Employees Table for EDA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Maximum date availiable in the Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------------------------------------+\n",
      "|greatest(max(last_date), max(hire_date))|\n",
      "+----------------------------------------+\n",
      "|                              2013-03-10|\n",
      "+----------------------------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "maxDateInDataset = employees.select(greatest(max(col('last_date')),max(col('hire_date'))))\n",
    "maxDateInDataset.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>emp_no</th>\n",
       "      <th>emp_title_id</th>\n",
       "      <th>birth_date</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>sex</th>\n",
       "      <th>hire_date</th>\n",
       "      <th>no_of_projects</th>\n",
       "      <th>last_performance_rating</th>\n",
       "      <th>left2</th>\n",
       "      <th>last_date</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure_Years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10001</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-09-02</td>\n",
       "      <td>Georgi</td>\n",
       "      <td>Facello</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-06-26</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "      <td>1994-07-30</td>\n",
       "      <td>60</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10002</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1964-06-02</td>\n",
       "      <td>Bezalel</td>\n",
       "      <td>Simmel</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-11-21</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>49</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10003</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1959-12-03</td>\n",
       "      <td>Parto</td>\n",
       "      <td>Bamford</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-08-28</td>\n",
       "      <td>1</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>54</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10004</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1954-05-01</td>\n",
       "      <td>Chirstian</td>\n",
       "      <td>Koblick</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-12-01</td>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>59</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10005</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1955-01-21</td>\n",
       "      <td>Kyoichi</td>\n",
       "      <td>Maliniak</td>\n",
       "      <td>M</td>\n",
       "      <td>1989-09-12</td>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>58</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>10006</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-04-20</td>\n",
       "      <td>Anneke</td>\n",
       "      <td>Preusig</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-06-02</td>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>60</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10007</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1957-05-23</td>\n",
       "      <td>Tzvetan</td>\n",
       "      <td>Zielinski</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-02-10</td>\n",
       "      <td>6</td>\n",
       "      <td>B</td>\n",
       "      <td>1</td>\n",
       "      <td>2002-09-18</td>\n",
       "      <td>56</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10008</td>\n",
       "      <td>e0001</td>\n",
       "      <td>1958-02-19</td>\n",
       "      <td>Saniya</td>\n",
       "      <td>Kalloufi</td>\n",
       "      <td>M</td>\n",
       "      <td>1994-09-15</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>55</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>10009</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1952-04-19</td>\n",
       "      <td>Sumant</td>\n",
       "      <td>Peac</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-02-18</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>61</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10010</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1963-06-01</td>\n",
       "      <td>Duangkaew</td>\n",
       "      <td>Piveteau</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-08-24</td>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>50</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10011</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1953-11-07</td>\n",
       "      <td>Mary</td>\n",
       "      <td>Sluis</td>\n",
       "      <td>F</td>\n",
       "      <td>1990-01-22</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>60</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>10012</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1960-10-04</td>\n",
       "      <td>Patricio</td>\n",
       "      <td>Bridgland</td>\n",
       "      <td>M</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>3</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>53</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>10013</td>\n",
       "      <td>s0002</td>\n",
       "      <td>1963-06-07</td>\n",
       "      <td>Eberhardt</td>\n",
       "      <td>Terkki</td>\n",
       "      <td>M</td>\n",
       "      <td>1985-10-20</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>50</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>10014</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1956-02-12</td>\n",
       "      <td>Berni</td>\n",
       "      <td>Genin</td>\n",
       "      <td>M</td>\n",
       "      <td>1987-03-11</td>\n",
       "      <td>9</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>57</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>10015</td>\n",
       "      <td>s0002</td>\n",
       "      <td>1959-08-19</td>\n",
       "      <td>Guoxiang</td>\n",
       "      <td>Nooteboom</td>\n",
       "      <td>M</td>\n",
       "      <td>1987-07-02</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>54</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10016</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1961-05-02</td>\n",
       "      <td>Kazuhito</td>\n",
       "      <td>Cappelletti</td>\n",
       "      <td>M</td>\n",
       "      <td>1995-01-27</td>\n",
       "      <td>5</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>52</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>10017</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1958-07-06</td>\n",
       "      <td>Cristinel</td>\n",
       "      <td>Bouloucos</td>\n",
       "      <td>F</td>\n",
       "      <td>1993-08-03</td>\n",
       "      <td>10</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>55</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10018</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1954-06-19</td>\n",
       "      <td>Kazuhide</td>\n",
       "      <td>Peha</td>\n",
       "      <td>F</td>\n",
       "      <td>1987-04-03</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>59</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10019</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1953-01-23</td>\n",
       "      <td>Lillian</td>\n",
       "      <td>Haddadi</td>\n",
       "      <td>M</td>\n",
       "      <td>1999-04-30</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>60</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10020</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1952-12-24</td>\n",
       "      <td>Mayuko</td>\n",
       "      <td>Warwick</td>\n",
       "      <td>M</td>\n",
       "      <td>1991-01-26</td>\n",
       "      <td>7</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>61</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    emp_no emp_title_id  birth_date first_name    last_name sex   hire_date  \\\n",
       "0    10001        e0003  1953-09-02     Georgi      Facello   M  1986-06-26   \n",
       "1    10002        s0001  1964-06-02    Bezalel       Simmel   F  1985-11-21   \n",
       "2    10003        e0003  1959-12-03      Parto      Bamford   M  1986-08-28   \n",
       "3    10004        e0003  1954-05-01  Chirstian      Koblick   M  1986-12-01   \n",
       "4    10005        s0001  1955-01-21    Kyoichi     Maliniak   M  1989-09-12   \n",
       "5    10006        e0003  1953-04-20     Anneke      Preusig   F  1989-06-02   \n",
       "6    10007        s0001  1957-05-23    Tzvetan    Zielinski   F  1989-02-10   \n",
       "7    10008        e0001  1958-02-19     Saniya     Kalloufi   M  1994-09-15   \n",
       "8    10009        e0003  1952-04-19     Sumant         Peac   F  1985-02-18   \n",
       "9    10010        e0002  1963-06-01  Duangkaew     Piveteau   F  1989-08-24   \n",
       "10   10011        s0001  1953-11-07       Mary        Sluis   F  1990-01-22   \n",
       "11   10012        e0003  1960-10-04   Patricio    Bridgland   M  1992-12-18   \n",
       "12   10013        s0002  1963-06-07  Eberhardt       Terkki   M  1985-10-20   \n",
       "13   10014        e0002  1956-02-12      Berni        Genin   M  1987-03-11   \n",
       "14   10015        s0002  1959-08-19   Guoxiang    Nooteboom   M  1987-07-02   \n",
       "15   10016        s0001  1961-05-02   Kazuhito  Cappelletti   M  1995-01-27   \n",
       "16   10017        s0001  1958-07-06  Cristinel    Bouloucos   F  1993-08-03   \n",
       "17   10018        e0003  1954-06-19   Kazuhide         Peha   F  1987-04-03   \n",
       "18   10019        s0001  1953-01-23    Lillian      Haddadi   M  1999-04-30   \n",
       "19   10020        e0002  1952-12-24     Mayuko      Warwick   M  1991-01-26   \n",
       "\n",
       "    no_of_projects last_performance_rating  left2   last_date  Age  \\\n",
       "0                9                       C      1  1994-07-30   60   \n",
       "1                8                       B      0        None   49   \n",
       "2                1                       C      0        None   54   \n",
       "3                5                       A      0        None   59   \n",
       "4                6                       A      0        None   58   \n",
       "5               10                       B      0        None   60   \n",
       "6                6                       B      1  2002-09-18   56   \n",
       "7                9                       C      0        None   55   \n",
       "8                8                       B      0        None   61   \n",
       "9                4                       A      0        None   50   \n",
       "10               6                       C      0        None   60   \n",
       "11               3                       A      0        None   53   \n",
       "12               6                       C      0        None   50   \n",
       "13               9                       B      0        None   57   \n",
       "14               4                       B      0        None   54   \n",
       "15               5                       C      0        None   52   \n",
       "16              10                       A      0        None   55   \n",
       "17               4                       B      0        None   59   \n",
       "18               4                       B      0        None   60   \n",
       "19               7                       B      0        None   61   \n",
       "\n",
       "    Tenure_Years  \n",
       "0              8  \n",
       "1             28  \n",
       "2             27  \n",
       "3             27  \n",
       "4             24  \n",
       "5             24  \n",
       "6             13  \n",
       "7             19  \n",
       "8             28  \n",
       "9             24  \n",
       "10            23  \n",
       "11            21  \n",
       "12            28  \n",
       "13            26  \n",
       "14            26  \n",
       "15            18  \n",
       "16            20  \n",
       "17            26  \n",
       "18            14  \n",
       "19            22  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "maxDateInDataset = employees.select(greatest(max(col('last_date')),max(col('hire_date')))).toPandas().loc[0][0]\n",
    "employees =  employees.withColumn('Age', maxDateInDataset.year-year(employees.birth_date))\\\n",
    "                .withColumn('Tenure_Years', F.when(col('left2')==1, year(employees.last_date)-year(employees.hire_date))\\\n",
    "                                                  .otherwise(maxDateInDataset.year-year(employees.hire_date)))\n",
    "employees.createOrReplaceTempView('employees_at')\n",
    "employees.toPandas().head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Exploratory Data Analysis**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **1. A list showing employee number, last name, first name, sex, and salary for each employee**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+---+------+\n",
      "|emp_no|  last_name|first_name|sex|salary|\n",
      "+------+-----------+----------+---+------+\n",
      "| 10001|    Facello|    Georgi|  M| 60117|\n",
      "| 10002|     Simmel|   Bezalel|  F| 65828|\n",
      "| 10003|    Bamford|     Parto|  M| 40006|\n",
      "| 10004|    Koblick| Chirstian|  M| 40054|\n",
      "| 10005|   Maliniak|   Kyoichi|  M| 78228|\n",
      "| 10006|    Preusig|    Anneke|  F| 40000|\n",
      "| 10007|  Zielinski|   Tzvetan|  F| 56724|\n",
      "| 10008|   Kalloufi|    Saniya|  M| 46671|\n",
      "| 10009|       Peac|    Sumant|  F| 60929|\n",
      "| 10010|   Piveteau| Duangkaew|  F| 72488|\n",
      "| 10011|      Sluis|      Mary|  F| 42365|\n",
      "| 10012|  Bridgland|  Patricio|  M| 40000|\n",
      "| 10013|     Terkki| Eberhardt|  M| 40000|\n",
      "| 10014|      Genin|     Berni|  M| 46168|\n",
      "| 10015|  Nooteboom|  Guoxiang|  M| 40000|\n",
      "| 10016|Cappelletti|  Kazuhito|  M| 70889|\n",
      "| 10017|  Bouloucos| Cristinel|  F| 71380|\n",
      "| 10018|       Peha|  Kazuhide|  F| 55881|\n",
      "| 10019|    Haddadi|   Lillian|  M| 44276|\n",
      "| 10020|    Warwick|    Mayuko|  M| 40000|\n",
      "+------+-----------+----------+---+------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT e.emp_no, last_name, first_name, sex, salary \n",
    "            FROM employees e\n",
    "            JOIN salaries s  ON e.emp_no = s.emp_no\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+---+------+\n",
      "|emp_no|  last_name|first_name|sex|salary|\n",
      "+------+-----------+----------+---+------+\n",
      "| 10001|    Facello|    Georgi|  M| 60117|\n",
      "| 10002|     Simmel|   Bezalel|  F| 65828|\n",
      "| 10003|    Bamford|     Parto|  M| 40006|\n",
      "| 10004|    Koblick| Chirstian|  M| 40054|\n",
      "| 10005|   Maliniak|   Kyoichi|  M| 78228|\n",
      "| 10006|    Preusig|    Anneke|  F| 40000|\n",
      "| 10007|  Zielinski|   Tzvetan|  F| 56724|\n",
      "| 10008|   Kalloufi|    Saniya|  M| 46671|\n",
      "| 10009|       Peac|    Sumant|  F| 60929|\n",
      "| 10010|   Piveteau| Duangkaew|  F| 72488|\n",
      "| 10011|      Sluis|      Mary|  F| 42365|\n",
      "| 10012|  Bridgland|  Patricio|  M| 40000|\n",
      "| 10013|     Terkki| Eberhardt|  M| 40000|\n",
      "| 10014|      Genin|     Berni|  M| 46168|\n",
      "| 10015|  Nooteboom|  Guoxiang|  M| 40000|\n",
      "| 10016|Cappelletti|  Kazuhito|  M| 70889|\n",
      "| 10017|  Bouloucos| Cristinel|  F| 71380|\n",
      "| 10018|       Peha|  Kazuhide|  F| 55881|\n",
      "| 10019|    Haddadi|   Lillian|  M| 44276|\n",
      "| 10020|    Warwick|    Mayuko|  M| 40000|\n",
      "+------+-----------+----------+---+------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees.join(salaries, on='emp_no').select('emp_no','last_name','first_name','sex','salary').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **2. A list showing first name, last name, and hire date for employees who were hired in 1986.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------+------------+----------+\n",
      "|first_name|   last_name| hire_date|\n",
      "+----------+------------+----------+\n",
      "|    Georgi|     Facello|1986-06-26|\n",
      "|     Parto|     Bamford|1986-08-28|\n",
      "| Chirstian|     Koblick|1986-12-01|\n",
      "|    Sanjiv|    Zschoche|1986-02-04|\n",
      "|      Kwee|    Schusler|1986-02-26|\n",
      "|   Kshitij|        Gils|1986-03-27|\n",
      "|  Zhongwei|       Rosen|1986-10-30|\n",
      "|   Xinglin|     Eugenio|1986-09-08|\n",
      "| Sudharsan|Flasterstein|1986-08-12|\n",
      "|    Kendra|     Hofting|1986-03-14|\n",
      "|    Hilari|      Morton|1986-07-15|\n",
      "|     Akemi|       Birch|1986-12-02|\n",
      "|    Lunjin|      Giveon|1986-10-02|\n",
      "|    Xuejia|      Ullian|1986-08-22|\n",
      "|   Chikara|    Rissland|1986-01-23|\n",
      "|  Domenick|    Peltason|1986-03-14|\n",
      "|    Zissis|    Pintelas|1986-02-11|\n",
      "|     Perry|   Shimshoni|1986-09-18|\n",
      "|  Kazuhito| Encarnacion|1986-08-21|\n",
      "|   Xiadong|       Perry|1986-11-05|\n",
      "+----------+------------+----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT first_name, last_name, hire_date\n",
    "            FROM employees\n",
    "            WHERE year(hire_date) = 1986\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------+------------+----------+\n",
      "|first_name|   last_name| hire_date|\n",
      "+----------+------------+----------+\n",
      "|    Georgi|     Facello|1986-06-26|\n",
      "|     Parto|     Bamford|1986-08-28|\n",
      "| Chirstian|     Koblick|1986-12-01|\n",
      "|    Sanjiv|    Zschoche|1986-02-04|\n",
      "|      Kwee|    Schusler|1986-02-26|\n",
      "|   Kshitij|        Gils|1986-03-27|\n",
      "|  Zhongwei|       Rosen|1986-10-30|\n",
      "|   Xinglin|     Eugenio|1986-09-08|\n",
      "| Sudharsan|Flasterstein|1986-08-12|\n",
      "|    Kendra|     Hofting|1986-03-14|\n",
      "|    Hilari|      Morton|1986-07-15|\n",
      "|     Akemi|       Birch|1986-12-02|\n",
      "|    Lunjin|      Giveon|1986-10-02|\n",
      "|    Xuejia|      Ullian|1986-08-22|\n",
      "|   Chikara|    Rissland|1986-01-23|\n",
      "|  Domenick|    Peltason|1986-03-14|\n",
      "|    Zissis|    Pintelas|1986-02-11|\n",
      "|     Perry|   Shimshoni|1986-09-18|\n",
      "|  Kazuhito| Encarnacion|1986-08-21|\n",
      "|   Xiadong|       Perry|1986-11-05|\n",
      "+----------+------------+----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees[year(col('hire_date'))==1986].select('first_name','last_name','hire_date').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **3. A list showing the manager of each department with the following information: department number, department name, the manager's employee number, last name, first name.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------+------------------+------+------------+-----------+-------+\n",
      "|dept_no|         dept_name|emp_no|   last_name| first_name|  title|\n",
      "+-------+------------------+------+------------+-----------+-------+\n",
      "|   d001|         Marketing|110022|  Markovitch|  Margareta|Manager|\n",
      "|   d001|         Marketing|110039|    Minakawa|   Vishwani|Manager|\n",
      "|   d002|           Finance|110085|       Alpin|       Ebru|Manager|\n",
      "|   d002|           Finance|110114|  Legleitner|      Isamu|Manager|\n",
      "|   d003|   Human Resources|110183|Ossenbruggen|    Shirish|Manager|\n",
      "|   d003|   Human Resources|110228|     Sigstam|    Karsten|Manager|\n",
      "|   d004|        Production|110303|     Wegerle|  Krassimir|Manager|\n",
      "|   d004|        Production|110344|       Cools|     Rosine|Manager|\n",
      "|   d004|        Production|110386|      Kieras|       Shem|Manager|\n",
      "|   d004|        Production|110420|    Ghazalie|      Oscar|Manager|\n",
      "|   d005|       development|110511|    Hagimont|   DeForest|Manager|\n",
      "|   d005|       development|110567|    DasSarma|       Leon|Manager|\n",
      "|   d006|Quality Management|110725|     Onuegbe|  Peternela|Manager|\n",
      "|   d006|Quality Management|110765|     Hofmeyr|     Rutger|Manager|\n",
      "|   d006|Quality Management|110800|     Quadeer|     Sanjoy|Manager|\n",
      "|   d006|Quality Management|110854|       Pesch|       Dung|Manager|\n",
      "|   d007|             Sales|111035|   Kaelbling|Przemyslawa|Manager|\n",
      "|   d007|             Sales|111133|       Zhang|      Hauke|Manager|\n",
      "|   d008|          Research|111400|     Staelin|       Arie|Manager|\n",
      "|   d008|          Research|111534|      Kambil|     Hilary|Manager|\n",
      "+-------+------------------+------+------------+-----------+-------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT dm.dept_no, d.dept_name, e.emp_no, e.last_name, e.first_name, t.title\n",
    "            FROM dept_manager dm \n",
    "            JOIN departments d ON dm.dept_no = d.dept_no\n",
    "            JOIN employees e ON dm.emp_no = e.emp_no\n",
    "            JOIN titles t ON e.emp_title_id = t.title_id AND t.title = 'Manager'\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------+------------------+------+------------+-----------+-------+\n",
      "|dept_no|         dept_name|emp_no|   last_name| first_name|  title|\n",
      "+-------+------------------+------+------------+-----------+-------+\n",
      "|   d001|         Marketing|110022|  Markovitch|  Margareta|Manager|\n",
      "|   d001|         Marketing|110039|    Minakawa|   Vishwani|Manager|\n",
      "|   d002|           Finance|110085|       Alpin|       Ebru|Manager|\n",
      "|   d002|           Finance|110114|  Legleitner|      Isamu|Manager|\n",
      "|   d003|   Human Resources|110183|Ossenbruggen|    Shirish|Manager|\n",
      "|   d003|   Human Resources|110228|     Sigstam|    Karsten|Manager|\n",
      "|   d004|        Production|110303|     Wegerle|  Krassimir|Manager|\n",
      "|   d004|        Production|110344|       Cools|     Rosine|Manager|\n",
      "|   d004|        Production|110386|      Kieras|       Shem|Manager|\n",
      "|   d004|        Production|110420|    Ghazalie|      Oscar|Manager|\n",
      "|   d005|       development|110511|    Hagimont|   DeForest|Manager|\n",
      "|   d005|       development|110567|    DasSarma|       Leon|Manager|\n",
      "|   d006|Quality Management|110725|     Onuegbe|  Peternela|Manager|\n",
      "|   d006|Quality Management|110765|     Hofmeyr|     Rutger|Manager|\n",
      "|   d006|Quality Management|110800|     Quadeer|     Sanjoy|Manager|\n",
      "|   d006|Quality Management|110854|       Pesch|       Dung|Manager|\n",
      "|   d007|             Sales|111035|   Kaelbling|Przemyslawa|Manager|\n",
      "|   d007|             Sales|111133|       Zhang|      Hauke|Manager|\n",
      "|   d008|          Research|111400|     Staelin|       Arie|Manager|\n",
      "|   d008|          Research|111534|      Kambil|     Hilary|Manager|\n",
      "+-------+------------------+------+------------+-----------+-------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "dept_manager.join(departments, on='dept_no')\\\n",
    "    .join(employees, on='emp_no')\\\n",
    "    .join(titles, employees.emp_title_id == titles.title_id)\\\n",
    "    .select('dept_no','dept_name','emp_no','last_name','first_name','title').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **4. A list showing the department of each employee with the following information: employee number, last name, first name, and department name.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+------------------+\n",
      "|emp_no|  last_name|first_name|         dept_name|\n",
      "+------+-----------+----------+------------------+\n",
      "| 10001|    Facello|    Georgi|       development|\n",
      "| 10002|     Simmel|   Bezalel|             Sales|\n",
      "| 10003|    Bamford|     Parto|        Production|\n",
      "| 10004|    Koblick| Chirstian|        Production|\n",
      "| 10005|   Maliniak|   Kyoichi|   Human Resources|\n",
      "| 10006|    Preusig|    Anneke|       development|\n",
      "| 10007|  Zielinski|   Tzvetan|          Research|\n",
      "| 10008|   Kalloufi|    Saniya|       development|\n",
      "| 10009|       Peac|    Sumant|Quality Management|\n",
      "| 10010|   Piveteau| Duangkaew|Quality Management|\n",
      "| 10011|      Sluis|      Mary|  Customer Service|\n",
      "| 10012|  Bridgland|  Patricio|       development|\n",
      "| 10013|     Terkki| Eberhardt|   Human Resources|\n",
      "| 10014|      Genin|     Berni|       development|\n",
      "| 10015|  Nooteboom|  Guoxiang|          Research|\n",
      "| 10016|Cappelletti|  Kazuhito|             Sales|\n",
      "| 10017|  Bouloucos| Cristinel|         Marketing|\n",
      "| 10018|       Peha|  Kazuhide|       development|\n",
      "| 10019|    Haddadi|   Lillian|          Research|\n",
      "| 10020|    Warwick|    Mayuko|        Production|\n",
      "+------+-----------+----------+------------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT e.emp_no, last_name, first_name, d.dept_name \n",
    "            FROM employees e \n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no \n",
    "            JOIN departments d ON de.dept_no = d.dept_no\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+------------------+\n",
      "|emp_no|  last_name|first_name|         dept_name|\n",
      "+------+-----------+----------+------------------+\n",
      "| 10001|    Facello|    Georgi|       development|\n",
      "| 10002|     Simmel|   Bezalel|             Sales|\n",
      "| 10003|    Bamford|     Parto|        Production|\n",
      "| 10004|    Koblick| Chirstian|        Production|\n",
      "| 10005|   Maliniak|   Kyoichi|   Human Resources|\n",
      "| 10006|    Preusig|    Anneke|       development|\n",
      "| 10007|  Zielinski|   Tzvetan|          Research|\n",
      "| 10008|   Kalloufi|    Saniya|       development|\n",
      "| 10009|       Peac|    Sumant|Quality Management|\n",
      "| 10010|   Piveteau| Duangkaew|Quality Management|\n",
      "| 10011|      Sluis|      Mary|  Customer Service|\n",
      "| 10012|  Bridgland|  Patricio|       development|\n",
      "| 10013|     Terkki| Eberhardt|   Human Resources|\n",
      "| 10014|      Genin|     Berni|       development|\n",
      "| 10015|  Nooteboom|  Guoxiang|          Research|\n",
      "| 10016|Cappelletti|  Kazuhito|             Sales|\n",
      "| 10017|  Bouloucos| Cristinel|         Marketing|\n",
      "| 10018|       Peha|  Kazuhide|       development|\n",
      "| 10019|    Haddadi|   Lillian|          Research|\n",
      "| 10020|    Warwick|    Mayuko|        Production|\n",
      "+------+-----------+----------+------------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees.join(dept_emp, on='emp_no')\\\n",
    "    .join(departments, on='dept_no')\\\n",
    "    .select('emp_no','last_name','first_name','dept_name').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **5. A list showing first name, last name, and sex for employees whose first name is \"Hercules\" and last names begin with \"B\"**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------+-------------+---+\n",
      "|first_name|    last_name|sex|\n",
      "+----------+-------------+---+\n",
      "|  Hercules|   Benzmuller|  M|\n",
      "|  Hercules|      Brendel|  F|\n",
      "|  Hercules|   Baranowski|  M|\n",
      "|  Hercules|     Barreiro|  M|\n",
      "|  Hercules|         Baer|  M|\n",
      "|  Hercules|Bernardinello|  F|\n",
      "|  Hercules|      Basagni|  M|\n",
      "|  Hercules|        Biran|  F|\n",
      "|  Hercules|    Bernatsky|  M|\n",
      "|  Hercules|         Bail|  F|\n",
      "|  Hercules|        Birge|  F|\n",
      "|  Hercules|      Bisiani|  F|\n",
      "|  Hercules|       Bodoff|  M|\n",
      "|  Hercules|        Biron|  F|\n",
      "|  Hercules|      Buchter|  M|\n",
      "|  Hercules|         Bain|  F|\n",
      "|  Hercules|         Bahr|  M|\n",
      "|  Hercules|         Baak|  M|\n",
      "|  Hercules|     Benantar|  F|\n",
      "|  Hercules|      Berstel|  F|\n",
      "+----------+-------------+---+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT first_name, last_name, sex \n",
    "            FROM employees \n",
    "            WHERE first_name = 'Hercules' AND last_name LIKE 'B%'\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------+-------------+---+\n",
      "|first_name|    last_name|sex|\n",
      "+----------+-------------+---+\n",
      "|  Hercules|   Benzmuller|  M|\n",
      "|  Hercules|      Brendel|  F|\n",
      "|  Hercules|   Baranowski|  M|\n",
      "|  Hercules|     Barreiro|  M|\n",
      "|  Hercules|         Baer|  M|\n",
      "|  Hercules|Bernardinello|  F|\n",
      "|  Hercules|      Basagni|  M|\n",
      "|  Hercules|        Biran|  F|\n",
      "|  Hercules|    Bernatsky|  M|\n",
      "|  Hercules|         Bail|  F|\n",
      "|  Hercules|        Birge|  F|\n",
      "|  Hercules|      Bisiani|  F|\n",
      "|  Hercules|       Bodoff|  M|\n",
      "|  Hercules|        Biron|  F|\n",
      "|  Hercules|      Buchter|  M|\n",
      "|  Hercules|         Bain|  F|\n",
      "|  Hercules|         Bahr|  M|\n",
      "|  Hercules|         Baak|  M|\n",
      "|  Hercules|     Benantar|  F|\n",
      "|  Hercules|      Berstel|  F|\n",
      "+----------+-------------+---+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees[(col('first_name') == 'Hercules') & (col('last_name').like('B%'))]\\\n",
    "    .select('first_name','last_name','sex').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **6. A list showing all employees in the Sales department, including their employee number, last name, first name, and department name.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+------------+----------+---------+\n",
      "|emp_no|   last_name|first_name|dept_name|\n",
      "+------+------------+----------+---------+\n",
      "| 10002|      Simmel|   Bezalel|    Sales|\n",
      "| 10016| Cappelletti|  Kazuhito|    Sales|\n",
      "| 10034|        Swan|     Bader|    Sales|\n",
      "| 10041|      Lenart|       Uri|    Sales|\n",
      "| 10050|      Dredge|   Yinghua|    Sales|\n",
      "| 10053|    Zschoche|    Sanjiv|    Sales|\n",
      "| 10061|      Herber|       Tse|    Sales|\n",
      "| 10068|     Brattka|  Charlene|    Sales|\n",
      "| 10087|     Eugenio|   Xinglin|    Sales|\n",
      "| 10089|Flasterstein| Sudharsan|    Sales|\n",
      "| 10093|     Desikan|   Sailaja|    Sales|\n",
      "| 10095|      Morton|    Hilari|    Sales|\n",
      "| 10099|     Sullins|    Valter|    Sales|\n",
      "| 10101|      Heyers|     Perla|    Sales|\n",
      "| 10107|        Baca|      Dung|    Sales|\n",
      "| 10125|     Hiltgen|     Syozo|    Sales|\n",
      "| 10136|    Pintelas|    Zissis|    Sales|\n",
      "| 10148|       Azumi|    Douadi|    Sales|\n",
      "| 10149|       Perry|   Xiadong|    Sales|\n",
      "| 10151|    Lichtner|   Itzchak|    Sales|\n",
      "+------+------------+----------+---------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT e.emp_no, e.last_name, e.first_name, d.dept_name\n",
    "            FROM employees e \n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no \n",
    "            JOIN departments d ON de.dept_no = d.dept_no AND d.dept_name = 'Sales'\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+------------+----------+---------+\n",
      "|emp_no|   last_name|first_name|dept_name|\n",
      "+------+------------+----------+---------+\n",
      "| 10002|      Simmel|   Bezalel|    Sales|\n",
      "| 10016| Cappelletti|  Kazuhito|    Sales|\n",
      "| 10034|        Swan|     Bader|    Sales|\n",
      "| 10041|      Lenart|       Uri|    Sales|\n",
      "| 10050|      Dredge|   Yinghua|    Sales|\n",
      "| 10053|    Zschoche|    Sanjiv|    Sales|\n",
      "| 10061|      Herber|       Tse|    Sales|\n",
      "| 10068|     Brattka|  Charlene|    Sales|\n",
      "| 10087|     Eugenio|   Xinglin|    Sales|\n",
      "| 10089|Flasterstein| Sudharsan|    Sales|\n",
      "| 10093|     Desikan|   Sailaja|    Sales|\n",
      "| 10095|      Morton|    Hilari|    Sales|\n",
      "| 10099|     Sullins|    Valter|    Sales|\n",
      "| 10101|      Heyers|     Perla|    Sales|\n",
      "| 10107|        Baca|      Dung|    Sales|\n",
      "| 10125|     Hiltgen|     Syozo|    Sales|\n",
      "| 10136|    Pintelas|    Zissis|    Sales|\n",
      "| 10148|       Azumi|    Douadi|    Sales|\n",
      "| 10149|       Perry|   Xiadong|    Sales|\n",
      "| 10151|    Lichtner|   Itzchak|    Sales|\n",
      "+------+------------+----------+---------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees.join(dept_emp, on='emp_no')\\\n",
    "    .join(departments[col('dept_name')=='Sales'], on='dept_no')\\\n",
    "    .select('emp_no','last_name','first_name','dept_name').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **7. A list showing all employees in the Sales and Development departments, including their employee number, last name, first name, and department name.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+-----------+\n",
      "|emp_no|  last_name|first_name|  dept_name|\n",
      "+------+-----------+----------+-----------+\n",
      "| 10001|    Facello|    Georgi|development|\n",
      "| 10002|     Simmel|   Bezalel|      Sales|\n",
      "| 10006|    Preusig|    Anneke|development|\n",
      "| 10008|   Kalloufi|    Saniya|development|\n",
      "| 10012|  Bridgland|  Patricio|development|\n",
      "| 10014|      Genin|     Berni|development|\n",
      "| 10016|Cappelletti|  Kazuhito|      Sales|\n",
      "| 10018|       Peha|  Kazuhide|development|\n",
      "| 10021|       Erde|     Ramzi|development|\n",
      "| 10022|     Famili|    Shahaf|development|\n",
      "| 10023| Montemayor|     Bojan|development|\n",
      "| 10025|     Heyers| Prasadram|development|\n",
      "| 10027|    Reistad|    Divier|development|\n",
      "| 10028|   Tempesti|  Domenick|development|\n",
      "| 10031|     Joslin|   Karsten|development|\n",
      "| 10034|       Swan|     Bader|      Sales|\n",
      "| 10037|   Makrucki|   Pradeep|development|\n",
      "| 10041|     Lenart|       Uri|      Sales|\n",
      "| 10043|    Tzvieli|    Yishay|development|\n",
      "| 10048|   Syrotiuk|   Florian|development|\n",
      "+------+-----------+----------+-----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT e.emp_no, e.last_name, e.first_name, d.dept_name\n",
    "            FROM employees e \n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no \n",
    "            JOIN departments d ON de.dept_no = d.dept_no AND d.dept_name IN ('Sales','development')\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-----------+----------+-----------+\n",
      "|emp_no|  last_name|first_name|  dept_name|\n",
      "+------+-----------+----------+-----------+\n",
      "| 10001|    Facello|    Georgi|development|\n",
      "| 10002|     Simmel|   Bezalel|      Sales|\n",
      "| 10006|    Preusig|    Anneke|development|\n",
      "| 10008|   Kalloufi|    Saniya|development|\n",
      "| 10012|  Bridgland|  Patricio|development|\n",
      "| 10014|      Genin|     Berni|development|\n",
      "| 10016|Cappelletti|  Kazuhito|      Sales|\n",
      "| 10018|       Peha|  Kazuhide|development|\n",
      "| 10021|       Erde|     Ramzi|development|\n",
      "| 10022|     Famili|    Shahaf|development|\n",
      "| 10023| Montemayor|     Bojan|development|\n",
      "| 10025|     Heyers| Prasadram|development|\n",
      "| 10027|    Reistad|    Divier|development|\n",
      "| 10028|   Tempesti|  Domenick|development|\n",
      "| 10031|     Joslin|   Karsten|development|\n",
      "| 10034|       Swan|     Bader|      Sales|\n",
      "| 10037|   Makrucki|   Pradeep|development|\n",
      "| 10041|     Lenart|       Uri|      Sales|\n",
      "| 10043|    Tzvieli|    Yishay|development|\n",
      "| 10048|   Syrotiuk|   Florian|development|\n",
      "+------+-----------+----------+-----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees.join(dept_emp, on='emp_no')\\\n",
    "    .join(departments[col('dept_name').isin('Sales','development')], on='dept_no')\\\n",
    "    .select('emp_no','last_name','first_name','dept_name').show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **8. A list showing the frequency count of employee last names, in descending order. ( i.e., how many employees share each last name**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-----------+-------------+\n",
      "|  last_name|Last_name_cnt|\n",
      "+-----------+-------------+\n",
      "|       Baba|          226|\n",
      "|     Gelosh|          223|\n",
      "|      Coorg|          223|\n",
      "|     Farris|          222|\n",
      "|    Sudbeck|          222|\n",
      "|     Adachi|          221|\n",
      "|     Osgood|          220|\n",
      "|     Neiman|          218|\n",
      "|    Mandell|          218|\n",
      "|     Masada|          218|\n",
      "|Boudaillier|          217|\n",
      "|    Wendorf|          217|\n",
      "|   Cummings|          216|\n",
      "|     Pettis|          216|\n",
      "|    Solares|          216|\n",
      "|     Mahnke|          216|\n",
      "|     Emmart|          215|\n",
      "|    Kulisch|          215|\n",
      "|   Birjandi|          215|\n",
      "| Maksimenko|          215|\n",
      "+-----------+-------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT last_name, count(*) AS Last_name_cnt\n",
    "            FROM employees\n",
    "            GROUP BY last_name\n",
    "            ORDER BY Last_name_cnt DESC\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-----------+-------------+\n",
      "|  last_name|Last_name_cnt|\n",
      "+-----------+-------------+\n",
      "|       Baba|          226|\n",
      "|     Gelosh|          223|\n",
      "|      Coorg|          223|\n",
      "|    Sudbeck|          222|\n",
      "|     Farris|          222|\n",
      "|     Adachi|          221|\n",
      "|     Osgood|          220|\n",
      "|     Masada|          218|\n",
      "|     Neiman|          218|\n",
      "|    Mandell|          218|\n",
      "|Boudaillier|          217|\n",
      "|    Wendorf|          217|\n",
      "|    Solares|          216|\n",
      "|   Cummings|          216|\n",
      "|     Pettis|          216|\n",
      "|     Mahnke|          216|\n",
      "|   Collette|          215|\n",
      "|     Emmart|          215|\n",
      "|   Birjandi|          215|\n",
      "|    Kulisch|          215|\n",
      "+-----------+-------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "employees.groupBy('last_name').agg(count('emp_no').alias('Last_name_cnt'))\\\n",
    "    .orderBy(col('Last_name_cnt').desc()).show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **9. Histogram to show the salary distribution among the employees**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (10, 6)})\n",
    "sns.distplot(salaries.select('salary').toPandas(), norm_hist=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoAAAAF5CAYAAAAPuJtvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3df3BU9b3/8ddmMfwQ0k1CgCQwUJhbiFJFEmVqvVeaCBFMCDpCuBEsv0SqCIoiiENikYIhMAIFDW29di6lpHrvHUJCSWgFW/FCL8GgYPjV8HOakJANCAESSPZ8//DrjqkQj8meXbLn+Zhxmj3v/fE+vFP2xeecPeswDMMQAAAAbCMk0A0AAADAvwiAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACb6RDoBtqj8+cvy+Ox7vKJkZFd5XbXWfb8sB4zDA7Msf1jhsGBObZOSIhD4eG337BGAGwFj8ewNAB+9Rpo35hhcGCO7R8zDA7M0bc4BAwAAGAzBEAAAACbIQACAADYDAEQAADAZgiAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQACAADYDAEQAADAZgiAAAAANtMh0A3gm65db1JUVLdAt+ET9Q2NunTxaqDbAAAAX0MAvAWF3uZU6ov5gW7DJwpWpulSoJsAAADNcAgYAADAZgiAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQACAADYDAEQAADAZgiAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQACAADYDAEQAADAZgiAAAAANuO3AJidna3ExEQNHDhQR48e/UZ97dq136jt379fY8aMUXJysqZOnSq3221pDQAAwA78FgCTkpK0ceNGxcbGfqP2+eefa//+/YqJifFuMwxD8+bNU2ZmpoqLi5WQkKAVK1ZYVgMAALALvwXAhIQERUdHf2P7tWvXtHjxYmVlZcnhcHi3HzhwQB07dlRCQoIkacKECSoqKrKsBgAAYBcBPwdw9erVGjNmjPr06dNse2VlZbMVwYiICHk8Hl24cMGSGgAAgF10COSLl5aW6sCBA3rppZcC2cZ3FhnZNdAttCtRUd0C3UJA2HW/gw1zbP+YYXBgjr4V0AC4d+9eHT9+XElJSZKks2fPatq0aVq2bJmio6NVUVHhvW9tba0cDodcLpclte/C7a6Tx2O0dre/VbD9kp87dynQLfhdVFQ3W+53sGGO7R8zDA7MsXVCQhw3XbQK6CHgGTNmaNeuXdqxY4d27NihXr166Z133tEDDzygwYMHq76+XiUlJZKkvLw8jRo1SpIsqQEAANiF31YAlyxZou3bt6umpkZTpkyRy+XS1q1bb3r/kJAQLV++XFlZWWpoaFBsbKxycnIsqwEAANiFwzAM645lBil/HAJOfTHfsuf3p4KVabZctudwRXBgju0fMwwOzLF1btlDwAAAAPA/AiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADbjtwCYnZ2txMREDRw4UEePHpUknT9/Xk899ZSSk5OVmpqqWbNmqba21vuY/fv3a8yYMUpOTtbUqVPldrstrQEAANiB3wJgUlKSNm7cqNjYWO82h8Oh6dOnq7i4WAUFBerTp49WrFghSTIMQ/PmzVNmZqaKi4uVkJBgaQ0AAMAu/BYAExISFB0d3Wyby+XSsGHDvLeHDBmiiooKSdKBAwfUsWNHJSQkSJImTJigoqIiy2oAAAB2ccucA+jxeLRp0yYlJiZKkiorKxUTE+OtR0REyOPx6MKFC5bUAAAA7KJDoBv4yuuvv64uXbpo4sSJgW7lW0VGdg10C+1KVFS3QLcQEHbd72DDHNs/ZhgcmKNv3RIBMDs7W6dOnVJubq5CQr5clIyOjvYeDpak2tpaORwOuVwuS2rfhdtdJ4/HaO3ufqtg+yU/d+5SoFvwu6iobrbc72DDHNs/ZhgcmGPrhIQ4brpoFfBDwG+++aYOHjyodevWKTQ01Lt98ODBqq+vV0lJiSQpLy9Po0aNsqwGAABgF35bAVyyZIm2b9+umpoaTZkyRS6XS6tWrVJubq769eunCRMmSJJ69+6tdevWKSQkRMuXL1dWVpYaGhoUGxurnJwcSbKkBgAAYBcOwzCsO5YZpPxxCDj1xXzLnt+fClam2XLZnsMVwYE5tn/MMDgwx9a5pQ8BAwAAwL8IgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAm2lVANyzZ4/27t3r614AAADgB6YC4MSJE7Vv3z5J0q9+9SvNnTtXc+fOVW5urqXNAQAAwPdMBcBjx45pyJAhkqT3339fGzZs0Hvvvae8vDxLmwMAAIDvdTBzJ4/HI4fDodOnT8swDA0YMECS9MUXX1jaHAAAAHzPVACMj4/X4sWLde7cOY0YMUKSdPr0aYWHh1vaHAAAAHzP1CHgZcuWKSwsTAMHDtRzzz0nSTp+/LiefPJJS5sDAACA75laAQwPD9fcuXObbRs+fLgV/QAAAMBiplYA3333XR06dEiStH//fg0fPlxJSUkqLS21tDkAAAD4nqkA+Nvf/la9e/eWJK1cuVKTJ0/WzJkztXTpUkubAwAAgO+ZCoCXLl1St27dVFdXpyNHjmjSpEkaN26cTpw4YXV/AAAA8DFT5wBGR0frk08+0d///nclJCTI6XSqrq5OTqfT6v4AAADgY6YC4Msvv6zZs2crNDRUa9askSTt3LlTP/zhDy1tDgAAAL5nKgA++OCD2rVrV7NtDz/8sB5++GFLmgIAAIB1TAVASSovL1dRUZHcbrcyMzN1+vRpXb9+XYMGDbKyPwAAAPiYqQ+BbNu2TU888YSqqqq0efNmSdKVK1f0xhtvWNocAAAAfM/UCuCaNWv07rvvKi4uTtu2bZMkDRo0SIcPH7a0OQAAAPieqRXA2tpa76Feh8Ph/d+vfgYAAED7YSoA3nnnncrPz2+2bevWrbrrrrtMvUh2drYSExM1cOBAHT161Lv9xIkTSk9PV3JystLT03Xy5MmA1QAAAOzCVAB89dVXtWrVKk2cOFFXrlzRtGnTtHr1ar3yyiumXiQpKUkbN25UbGxss+1ZWVnKyMhQcXGxMjIylJmZGbAaAACAXZgKgAMGDNC2bduUkZGh559/Xo899pgKCgrUr18/Uy+SkJCg6OjoZtvcbrfKysqUkpIiSUpJSVFZWZlqa2v9XgMAALATUx8C+eCDDzR8+HCNHj3aZy9cWVmpnj17er9NxOl0qkePHqqsrJRhGH6tRURE+Gy/AAAAbnWmAuDq1av16quvavTo0Ro7dqzpc/+CVWRk10C30K5ERXULdAsBYdf9DjbMsf1jhsGBOfqWqQC4ZcsWHT58WPn5+Zo1a5Y6d+6stLQ0jRkzRr17927VC0dHR6uqqkpNTU1yOp1qampSdXW1oqOjZRiGX2vfldtdJ4/HaNV+mxFsv+Tnzl0KdAt+FxXVzZb7HWyYY/vHDIMDc2ydkBDHTRetTJ0DKH153b/58+frL3/5i7KyslRUVKQRI0boiSee0JYtW+TxeL5TU5GRkYqLi1NhYaEkqbCwUHFxcYqIiPB7DQAAwE4chmGYXso6ffq0tmzZoi1btsjhcGjs2LGKjo7Wpk2bFBUVpbVr197wcUuWLNH27dtVU1Oj8PBwuVwubd26VeXl5VqwYIEuXryosLAwZWdnq3///pLk99p34Y8VwNQX87/9ju1Awco0W/6rjX+tBgfm2P4xw+DAHFunpRVAUwFw48aNys/P16lTp/Twww/r0Ucf1ZAhQ7z1q1ev6v7771dpaanvur6FEQDNIwCiPWOO7R8zDA7MsXVaCoCmzgH861//qilTpigpKUmhoaHfqHfu3Fm//OUv29YlAAAA/MJUAFy/fr0kyePxqLq6Wt27d1dISPPTBx944AHfdwcAAACfM/UhkLq6Or388su666679G//9m+66667NH/+fF26xHIsAABAe2MqAC5ZskRXr15VQUGBPvvsMxUUFOjq1atasmSJ1f0BAADAx0wdAv7oo4/05z//WZ07d5Ykff/739eyZcs0YsQIS5sDAACA75laAezYseM3vjP3/PnzN/xACAAAAG5tplYAH3/8cU2dOlWTJ09WTEyMKioq9Nvf/lbjx4+3uj8AAAD4mKkA+LOf/Uw9evRQYWGhqqur1aNHD02fPl2PP/641f0BAADAx0wFQIfDoccff5zABwAAEARuGgD/67/+y9QTEAoBAADal5sGwPz8b/8qsq9WBgEAANB+3DQAbtiwwZ99AAAAwE9MnQMoSRcvXtSHH37o/RDI8OHDFRYWZmVvAAAAsICp6wDu3r1biYmJ2rBhgw4cOKDf/e53SkxM1O7du63uDwAAAD5magXw9ddf1+LFizV69Gjvtm3btunnP/+5ioqKLGsOAAAAvmdqBbC6ulrJycnNto0YMUI1NTWWNAUAAADrmAqAY8eO1caNG5tt27Rpk8aOHWtJUwAAALCOqUPAn3/+uTZt2qTf/OY36tmzp6qqqlRbW6u77rpLTzzxhPd+/xwSAQAAcOsxFQDHjx/P9/4CAAAECVMB8NFHH7W6DwAAAPiJ6esAlpSUqKysTFeuXGm2febMmT5vCgAAANYxfRmYbdu2KSEhQR07dvRudzgcljUGAAAAa5gKgAUFBSooKFDPnj2t7gcAAAAWM3UZmF69eik0NNTqXgAAAOAHplYAf/GLX2jRokV65JFH1L1792a1e++915LGAAAAYA3T1wH861//qr1796pTp07e7Q6HQx9++KFVvQEAAMACpgLgm2++qdzcXN1///1W9wMAAACLmToHsHPnzkpISLC6FwAAAPiBqQA4e/ZsLV26VOfOnZPH42n2HwAAANoXU4eAFy5cKEn6wx/+4N1mGIYcDocOHTpkTWcAAACwhKkA+MEHH1jdBwAAAPzEVACMjY21ug8AAAD4SYvnAI4dO7bZ7VdeeaXZ7R/96Ec+aWLnzp0aO3as0tLSlJqaqu3bt0uSTpw4ofT0dCUnJys9PV0nT570PsaKGgAAgB20GABPnTrV7PY/Hwqur69vcwOGYejll1/W8uXLlZ+fr5ycHM2fP18ej0dZWVnKyMhQcXGxMjIylJmZ6X2cFTUAAAA7aDEAOhyOZrcNw2ix3uomQkJ06dIlSdKlS5fUo0cPnT9/XmVlZUpJSZEkpaSkqKysTLW1tXK73T6vAQAA2IWpcwC/4qvA98/PuWrVKj3zzDPq0qWLLl++rPXr16uyslI9e/aU0+mUJDmdTvXo0UOVlZUyDMPntYiICJ/vGwAAwK2oxQB47do1rV692nu7vr6+2e3r16+3uYHGxkatX79eb731luLj47Vv3z698MILWr58eZuf2yqRkV0D3UK7EhXVLdAtBIRd9zvYMMf2jxkGB+boWy0GwNTUVJ09e9Z7+5FHHml2+6tDqW1x6NAhVVdXKz4+XpIUHx+vzp07q2PHjqqqqlJTU5OcTqeamppUXV2t6OhoGYbh89p34XbXyeMxvv2OrRRsv+Tnzl0KdAt+FxXVzZb7HWyYY/vHDIMDc2ydkBDHTRetWgyAy5Yts6Shr+vVq5fOnj2r48ePq3///iovL1dNTY369u2ruLg4FRYWKi0tTYWFhYqLi/MeqrWiBgAAYAcO458/2REAW7Zs0a9//WvvOYazZ8/WQw89pPLyci1YsEAXL15UWFiYsrOz1b9/f0mypGaWP1YAU1/Mt+z5/algZZot/9XGv1aDA3Ns/5hhcGCOrdPSCuAtEQDbGwKgeQRAtGfMsf1jhsGBObZOSwGwxcvAAAAAIPjcNABmZ2d7f969e7dfmgEAAID1bhoA33vvPe/Pzz77rF+aAQAAgPVu+ingQYMGafbs2RowYMA3rgf4dXPmzLGsOQAAAPjeTQPgmjVr9Ic//EEVFRWS1Oz6fwAAAGi/bhoAIyMj9cwzz0iSmpqa/HJNQAAAAFjP1HcBL1u2TF988YV27typqqoq9ezZU8OHD5fL5bK6PwAAAPiYqcvAlJaWasSIEcrLy9ORI0eUl5enkSNHqrS01Or+AAAA4GOmVgCXLl2qrKwsPfLII95tf/zjH7VkyRL993//t2XNAQAAwPdMrQCePHlSo0aNarYtOTlZp0+ftqQpAAAAWMdUAOzbt6+2bt3abFtRUZH69OljSVMAAACwjqlDwAsXLtTMmTO1YcMGxcTE6B//+IdOnTql3Nxcq/sDAACAj5kKgEOHDtWf/vQnffjhh6qurtZPfvITPfjgg3wKGAAAoB0yFQAl6Xvf+57S0tKs7AUAAAB+YOocQAAAAAQPAiAAAIDNEAABAABsplUBcM+ePdq7d6+vewEAAIAfmAqAEydO1L59+yRJv/rVrzR37lzNnTuXy8AAAAC0Q6YC4LFjxzRkyBBJ0vvvv68NGzbovffeU15enqXNAQAAwPdMXQbG4/HI4XDo9OnTMgxDAwYMkCR98cUXljYHAAAA3zMVAOPj47V48WKdO3dOI0aMkCSdPn1a4eHhljYHAAAA3zN1CHjZsmUKCwvTwIED9dxzz0mSjh8/rieffNLS5gAAAOB7plYA9+zZo7lz5zbbNnz4cBUVFVnSFAAAAKxjagXw1VdfveH2zMxMnzYDAAAA67W4AnjmzBlJkmEY3p+/XgsNDbWuMwAAAFiixQA4YsQIORwOGYbh/fDHV7p37+49HxAAAADtR4sB8PDhw5K+vBD07373O780BAAAAGuZOgeQ8AcAABA8TH0K+MyZM1q1apUOHTqkK1euNKt9+OGHVvQFAAAAi5gKgC+99JL69Omj+fPnq3Pnzlb3BAAAAAuZCoDHjh3Tpk2bFBJi6ogxAAAAbmGmEt29996rsrIyy5poaGhQVlaWRo4cqdTUVC1atEiSdOLECaWnpys5OVnp6ek6efKk9zFW1AAAAOzA1ApgbGyspk2bppEjR6p79+7NanPmzGlzEzk5OerYsaOKi4vlcDhUU1MjScrKylJGRobS0tKUn5+vzMxM/ed//qdlNQAAADswtQJ49epVJSYmqrGxUWfPnm32X1tdvnxZmzdv1pw5c+RwOCR9eY1Bt9utsrIypaSkSJJSUlJUVlam2tpaS2oAAAB2YWoFcNmyZZY1cObMGblcLq1du1Z/+9vfdPvtt2vOnDnq1KmTevbsKafTKUlyOp3q0aOHKisrZRiGz2sRERGme46M7OrjP4XgFhXVLdAtBIRd9zvYMMf2jxkGB+boW6YvA3Mzffr0aVMDjY2NOnPmjO644w7Nnz9fn376qWbOnKnVq1e36Xmt5HbXyeMxLHv+YPslP3fuUqBb8LuoqG623O9gwxzbP2YYHJhj64SEOG66aGUqAH79K+G+8tXh2kOHDrWpuZiYGHXo0MF7WPbuu+9WeHi4OnXqpKqqKjU1NcnpdKqpqUnV1dWKjo6WYRg+rwEAANiFqXMADx8+rEOHDunw4cM6fPiwPvroI40fP17Lly9vcwMREREaNmyYPv74Y0lffkrX7XarX79+iouLU2FhoSSpsLBQcXFxioiIUGRkpM9rAAAAduEwvr6s9x1cu3ZNycnJ2rlzZ5ubOHPmjBYuXKgLFy6oQ4cOev755/Xggw+qvLxcCxYs0MWLFxUWFqbs7Gz1799fkiypmeWPQ8CpL+Zb9vz+VLAyzZbL9hyuCA7Msf1jhsGBObZOS4eAWx0ADx8+rMmTJ2vPnj1taq49IgCaRwBEe8Yc2z9mGByYY+u0+RzAjIwM7zl/0peXhfn73/+uZ5991jcdAgAAwG9MBcBx48Y1u925c2cNGjRI/fr1s6InAAAAWMhUAHz00Uet7gMAAAB+YupTwNevX9eaNWuUlJSkH/7wh0pKStKaNWt07do1q/sDAACAj5laAczJydFnn32mn//854qJiVFFRYXeeust1dXVaeHChVb3CAAAAB8yFQCLioqUn5+v8PBwSVL//v11xx13KC0tjQAIAADQzpg6BHyzK8W08goyAAAACCBTK4APP/ywfvazn+nZZ59VTEyM/vGPf+jtt9/WqFGjrO4P7dy1601B893G9Q2NunTxaqDbAACgzUwFwHnz5untt9/W4sWLVV1drZ49e2r06NF65plnrO4P7Vzobc6guqg1lyEFAAQDUwEwNDRUc+bM0Zw5c6zuBwAAABZr8RzAffv2KScn54a1FStWaP/+/ZY0BQAAAOu0GADXr1+ve++994a1e++9V7m5uZY0BQAAAOu0GAAPHTqkf/3Xf71h7cc//rEOHjxoSVMAAACwTosBsK6uTtevX79hrbGxUZcvX7akKQAAAFinxQDYv39/7dq164a1Xbt2qX///pY0BQAAAOu0GAAnT56srKwsbd++XR6PR5Lk8Xi0fft2vfbaa5oyZYpfmgQAAIDvtHgZmNTUVNXU1Gj+/Pm6fv26XC6XLly4oNDQUM2ePVspKSn+6hMAAAA+8q3XAZwyZYrGjRun0tJSXbhwQS6XS/fcc4+6du3qj/4AAADgY6YuBN21a9ebfhoYAAAA7UuL5wACAAAg+BAAAQAAbIYACAAAYDMEQAAAAJshAAIAANgMARAAAMBmCIAAAAA2QwAEAACwGQIgAACAzRAAAQAAbIYACAAAYDMEQAAAAJu5pQLg2rVrNXDgQB09elSStH//fo0ZM0bJycmaOnWq3G63975W1AAAAOzglgmAn3/+ufbv36+YmBhJkmEYmjdvnjIzM1VcXKyEhAStWLHCshoAAIBddAh0A5J07do1LV68WCtWrNBPf/pTSdKBAwfUsWNHJSQkSJImTJigpKQkLVu2zJIa8G2uXW9SVFQ30/f/Lvf1p/qGRl26eDXQbQAAAuiWCICrV6/WmDFj1KdPH++2yspK72qgJEVERMjj8ejChQuW1Fwul8V7ifYu9DanUl/MD3QbbVawMk2XAt0EACCgAh4AS0tLdeDAAb300kuBbsW0yMiugW4BaJNbdXXyVsSfVfvHDIMDc/StgAfAvXv36vjx40pKSpIknT17VtOmTdOkSZNUUVHhvV9tba0cDodcLpeio6N9Xvsu3O46eTxGa3f5W/FLDqudO8caoBlRUd34s2rnmGFwYI6tExLiuOmiVcA/BDJjxgzt2rVLO3bs0I4dO9SrVy+98847mj59uurr61VSUiJJysvL06hRoyRJgwcP9nkNAADALgK+AngzISEhWr58ubKystTQ0KDY2Fjl5ORYVgMAALCLWy4A7tixw/vz0KFDVVBQcMP7WVEDAACwg4AfAgYAAIB/EQABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2EyHQDcAwL+uXW9SVFS3QLfhE/UNjbp08Wqg2wCAdocACNhM6G1Opb6YH+g2fKJgZZouBboJAGiHOAQMAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALAZAiAAAIDNEAABAABshgAIAABgMwEPgOfPn9dTTz2l5ORkpaamatasWaqtrZUk7d+/X2PGjFFycrKmTp0qt9vtfZwVNQAAADsIeAB0OByaPn26iouLVVBQoD59+mjFihUyDEPz5s1TZmamiouLlZCQoBUrVkiSJTUAAAC7CHgAdLlcGjZsmPf2kCFDVFFRoQMHDqhjx45KSEiQJE2YMEFFRUWSZEkNAADALgIeAL/O4/Fo06ZNSkxMVGVlpWJiYry1iIgIeTweXbhwwZIaAACAXXQIdANf9/rrr6tLly6aOHGi/vSnPwW6nZuKjOwa6BYA/H9RUd3a9fPDeswwODBH37plAmB2drZOnTql3NxchYSEKDo6WhUVFd56bW2tHA6HXC6XJbXvwu2uk8djtGFvW8YvOWDeuXOXLHvuqKhulj4/rMcMgwNzbJ2QEMdNF61uiUPAb775pg4ePKh169YpNDRUkjR48GDV19erpKREkpSXl6dRo0ZZVgMAALCLgK8AHjt2TLm5uerXr58mTJggSerdu7fWrVun5cuXKysrSw0NDYqNjVVOTo4kKSQkxOc1AAAAuwh4APyXf/kXHTly5Ia1oUOHqqCgwG81AAAAO7glDgEDAADAfwiAAAAANkMABAAAsBkCIAAAgM0E/EMgANBa1643Bc2FoOsbGnXp4lW/vBYAEAABtFuhtzmV+mJ+oNvwiYKVaeIytwD8hUPAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQACAADYDAEQAADAZgiAAAAANkMABAAAsBkCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQACAADYDAEQAADAZjoEugEAgHTtepOioroFuo02q29o1KWLVwPdBoBvQQAEgFtA6G1Opb6YH+g22qxgZZouBboJAN+KQ8AAAAA2QwAEAACwGQIgAACAzRAAAQAAbIYACAAAYDMEQAAAAJshAAIAANgM1wEEAPjMrXhB69b2w0WtEcxsGQBPnDihBQsW6MKFC3K5XMrOzla/fv0C3RYAtHvBckFriYtaI7jZ8hBwVlaWMjIyVFxcrIyMDGVmZga6JQAAAL+xXQB0u90qKytTSkqKJCklJUVlZWWqra0NcGcAAAD+YbtDwJWVlerZs6ecTqckyel0qkePHqqsrFRERISp55AE9boAAAoESURBVAgJcVjZoiSpR3hny1/DX9iXW0+w7IfEvtyKgmU/bsXzGVuroaFRdXX1gW6jTfzx3htsWvozcxiGYfixl4A7ePCg5s+fr61bt3q3jR49Wjk5ObrzzjsD2BkAAIB/2O4QcHR0tKqqqtTU1CRJampqUnV1taKjowPcGQAAgH/YLgBGRkYqLi5OhYWFkqTCwkLFxcWZPvwLAADQ3tnuELAklZeXa8GCBbp48aLCwsKUnZ2t/v37B7otAAAAv7BlAAQAALAz2x0CBgAAsDsCIAAAgM0QAAEAAGyGAAgAAGAzBEAAAACbIQBaaO3atRo4cKCOHj0qSdq/f7/GjBmj5ORkTZ06VW6323tfK2pom4aGBmVlZWnkyJFKTU3VokWLJEknTpxQenq6kpOTlZ6erpMnT3ofY0UNrbdz506NHTtWaWlpSk1N1fbt2yUxw1tZdna2EhMTm/3dKfl/ZsyzbW40x/Pnz+upp55ScnKyUlNTNWvWLNXW1nofw3uknxmwxMGDB41p06YZw4cPN44cOWJ4PB7joYceMvbu3WsYhmGsW7fOWLBggWEYhiU1tN3rr79u/OIXvzA8Ho9hGIZx7tw5wzAMY9KkScbmzZsNwzCMzZs3G5MmTfI+xooaWsfj8RgJCQnGkSNHDMMwjEOHDhlDhgwxmpqamOEtbO/evUZFRYXxk5/8xDs7w/D//++YZ9vcaI7nz5839uzZ473PG2+8YbzyyiuGYVjzPsh7ZMsIgBZoaGgwxo8fb5w+fdr7y//pp58ajzzyiPc+brfbGDJkiGEYhiU1tE1dXZ0RHx9v1NXVNdteU1NjxMfHG42NjYZhGEZjY6MRHx9vuN1uS2poPY/HY9x3331GSUmJYRiG8X//93/GyJEjmWE78fXg4O+ZMU/f+ecg/3VFRUXGT3/6U8MwrHkf5D2yZR0CvQIZjFavXq0xY8aoT58+3m2VlZWKiYnx3o6IiJDH49GFCxcsqblcLov3MridOXNGLpdLa9eu1d/+9jfdfvvtmjNnjjp16qSePXvK6XRKkpxOp3r06KHKykoZhuHzGl9R2HoOh0OrVq3SM888oy5duujy5ctav369KisrmWE74++ZMU/reTwebdq0SYmJiZJ4jwwEzgH0sdLSUh04cEAZGRmBbgVt0NjYqDNnzuiOO+7Q//zP/+ill17Sc889pytXrgS6NZjU2Nio9evX66233tLOnTv19ttv64UXXmCGwC3g9ddfV5cuXTRx4sRAt2JbrAD62N69e3X8+HElJSVJks6ePatp06Zp0qRJqqio8N6vtrZWDodDLpdL0dHRPq+hbWJiYtShQwelpKRIku6++26Fh4erU6dOqqqqUlNTk5xOp5qamlRdXa3o6GgZhuHzGlrv0KFDqq6uVnx8vCQpPj5enTt3VseOHZlhOxMdHe3XmTFPa2VnZ+vUqVPKzc1VSMiX61BWvA/yHtkyVgB9bMaMGdq1a5d27NihHTt2qFevXnrnnXc0ffp01dfXq6SkRJKUl5enUaNGSZIGDx7s8xraJiIiQsOGDdPHH38s6ctPBLrdbvXr109xcXEqLCyUJBUWFiouLk4RERGKjIz0eQ2t16tXL509e1bHjx+XJJWXl6umpkZ9+/Zlhu2MFXNhnoHx5ptv6uDBg1q3bp1CQ0O92614H+Q9smUOwzCMQDcRzBITE5Wbm6sf/OAH+uSTT5SVlaWGhgbFxsYqJydH3bt3lyRLamibM2fOaOHChbpw4YI6dOig559/Xg8++KDKy8u1YMECXbx4UWFhYcrOzlb//v0lyZIaWm/Lli369a9/LYfDIUmaPXu2HnroIWZ4C1uyZIm2b9+umpoahYeHy+VyaevWrX6fGfNsmxvNcdWqVUpJSVG/fv3UqVMnSVLv3r21bt06Sda8D/IeeXMEQAAAAJvhEDAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAslpiYqP/93/9t9ePvuecenTlzxocdAbA7AiAAmFRSUqIJEyYoPj5e9913nyZMmKDPPvvM8tctLS1t9t3iANBWfBUcAJhQV1enmTNn6rXXXtOoUaN0/fp1lZSUNPs2A19rbGxUhw78NQ3A91gBBAATTpw4IUlKSUmR0+lUp06d9MADD2jQoEE6ffq0nnzySQ0bNkzDhg3Tiy++qIsXL97weT777DOlp6crISFBDzzwgBYvXqxr16556wMHDtTGjRs1cuRIjRw50rvt1KlTkqRr164pOztbw4cP1/3336/MzEzV19dL+vK7Tp9++mklJCTovvvuU0ZGhjwej5V/LADaKQIgAJjw/e9/X06nU/Pnz9df/vIXffHFF96aYRh6+umn9dFHH2nbtm06e/asfvnLX97weUJCQvTKK69oz549ysvL0+7du/X73/++2X3+/Oc/67333tMf//jHbzw+JydHJ06c0ObNm7V9+3ZVV1d7v0rr3XffVc+ePbV79259/PHHmjt3rvdr8ADg6wiAAGBC165d9fvf/14Oh0OLFi3Sj370I82cOVM1NTXq27evfvzjHys0NFQRERGaMmWK9u7de8PnGTx4sIYMGaIOHTqod+/eSk9P/8Z9Z8yYIZfL5f2+1K8YhqH3339fCxculMvlUteuXfX0009r69atkqQOHTro3Llzqqio0G233aaEhAQCIIAb4uQSADBpwIABeuONNyRJ5eXlmjdvnpYuXapXX31VS5YsUUlJiS5fvizDMBQWFnbD5zhx4oTeeOMNHTx4UFevXlVTU5PuvPPOZveJjo6+4WNra2t19epVPfbYY95thmF4D/NOmzZNa9eu1dSpUyVJ6enpmjFjRpv3G0DwYQUQAFphwIABeuyxx3Ts2DGtXLlSDodDW7Zs0SeffKKcnBwZhnHDx7322mvq37+/iouL9cknn+iFF174xn1vtmoXHh6uTp06aevWrSopKVFJSYn27dun0tJSSV+uUi5YsEAffPCBcnNz9e6772r37t2+3XEAQYEACAAmlJeX6z/+4z909uxZSVJlZaUKCwt199136/Lly+rSpYvCwsJUVVWl3/zmNzd9nsuXL+v222/X7bffrvLycm3atMl0DyEhIRo3bpyWLl0qt9stSaqqqtJHH30kSdq5c6dOnTolwzDUtWtXOZ1OhYTw1zyAb+JvBgAwoWvXrvr00081btw4DRkyROPHj9cPfvADLViwQLNmzVJZWZkSEhI0Y8YM76d3b2T+/PkqLCzU0KFDtWjRIo0ePfo79TFv3jz17dtX48eP19ChQzV58mTvJ5RPnTqlKVOm6J577lF6err+/d//XcOGDWvTfgMITg7jZscpAAAAEJRYAQQAALAZAiAAAIDNEAABAABshgAIAABgMwRAAAAAmyEAAgAA2AwBEAAAwGYIgAAAADZDAAQAALCZ/wcL6bhjGgVfqQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(salaries.select('salary').toPandas().values)\n",
    "plt.xlabel('Salaries')\n",
    "plt.ylabel('Counts of Employess')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **10. Bar graph to show the Average salary per title (designation)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>title</th>\n",
       "      <th>Avg_Salary</th>\n",
       "      <th>Max_Salary</th>\n",
       "      <th>Min_Salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Staff</td>\n",
       "      <td>58465.382850</td>\n",
       "      <td>129492</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Assistant Engineer</td>\n",
       "      <td>48564.434447</td>\n",
       "      <td>99683</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Senior Staff</td>\n",
       "      <td>58550.172704</td>\n",
       "      <td>127238</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>48506.799871</td>\n",
       "      <td>110449</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Engineer</td>\n",
       "      <td>48535.336511</td>\n",
       "      <td>100683</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Manager</td>\n",
       "      <td>51531.041667</td>\n",
       "      <td>72446</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Technique Leader</td>\n",
       "      <td>48582.896092</td>\n",
       "      <td>104065</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                title    Avg_Salary  Max_Salary  Min_Salary\n",
       "0               Staff  58465.382850      129492       40000\n",
       "1  Assistant Engineer  48564.434447       99683       40000\n",
       "2        Senior Staff  58550.172704      127238       40000\n",
       "3     Senior Engineer  48506.799871      110449       40000\n",
       "4            Engineer  48535.336511      100683       40000\n",
       "5             Manager  51531.041667       72446       40000\n",
       "6    Technique Leader  48582.896092      104065       40000"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q10 = spark.sql(\"\"\"\n",
    "            SELECT \n",
    "                t.title, \n",
    "                avg(Salary) Avg_Salary, \n",
    "                max(Salary) Max_Salary, \n",
    "                min(Salary) Min_Salary\n",
    "            FROM salaries s\n",
    "            JOIN employees e ON s.emp_no = e.emp_no\n",
    "            JOIN titles t ON e.emp_title_id = t.title_id\n",
    "            GROUP BY t.title\n",
    "         \"\"\").toPandas()\n",
    "q10.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (15, 6)})\n",
    "sns.barplot(data=q10, x='title', y='Avg_Salary')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **11. Calculate employee tenure & show the tenure distribution among the employees**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>Tenure_Years</th>\n",
       "      <th>Tenure_cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>27</td>\n",
       "      <td>32530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>31832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26</td>\n",
       "      <td>30175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>28210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24</td>\n",
       "      <td>25607</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Tenure_Years  Tenure_cnt\n",
       "0            27       32530\n",
       "1            28       31832\n",
       "2            26       30175\n",
       "3            25       28210\n",
       "4            24       25607"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q11 = spark.sql(\"\"\"\n",
    "           SELECT \n",
    "                CASE WHEN last_date IS NULL \n",
    "                     THEN (SELECT year(greatest(max(last_date), max(hire_date))) FROM employees)-year(hire_date)\n",
    "                ELSE year(last_date)-year(hire_date) END AS Tenure_Years, \n",
    "                count(*) AS Tenure_cnt\n",
    "            FROM employees\n",
    "            GROUP BY Tenure_Years\n",
    "            ORDER BY Tenure_cnt DESC\n",
    "         \"\"\").toPandas()\n",
    "q11.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>Tenure_Years</th>\n",
       "      <th>Tenure_cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>27</td>\n",
       "      <td>32530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>28</td>\n",
       "      <td>31832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26</td>\n",
       "      <td>30175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25</td>\n",
       "      <td>28210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24</td>\n",
       "      <td>25607</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Tenure_Years  Tenure_cnt\n",
       "0            27       32530\n",
       "1            28       31832\n",
       "2            26       30175\n",
       "3            25       28210\n",
       "4            24       25607"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "employees.groupBy('Tenure_Years').agg(count('emp_no').alias('Tenure_cnt'))\\\n",
    "    .orderBy(col('Tenure_cnt').desc()).toPandas().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA5AAAAF5CAYAAADpmMVHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3df1iUZcL+/3NmmAFRZAQVUdssN13KLJXW2rJaLM3N0iy/sXz7pZmlRmZZS/6ARM1FfTLdaGnLrZ5dV7ddf2tF7mq12ubmVpuGj5lZURIiiIKKwMz9+cPHeSRBL3HuGdT36zg6jrivmeu8wOGCk/ueGYdlWZYAAAAAADgJZ7gXAAAAAAA4M1AgAQAAAABGKJAAAAAAACMUSAAAAACAEQokAAAAAMAIBRIAAAAAYIQCCQAAAAAwEhHuBTRVe/cekN/PW2QCAAAAOLc4nQ61atW83jEKZAP8fosCCQAAAADH4BJWAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEYokAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGIkI9wKaurjYKLk8btvm91XXqGxflW3zAwAAAECwUCBPwuVxq+S3f7Rt/jaj7pJEgQQAAADQ9HEJKwAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAY4UV0mig7X/2VV34FAAAA0BgUyCbK5XFrd95ztszd9qFHxSu/AgAAADhVXMIKAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEYokAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEZCViBHjx6tW2+9VYMHD1ZaWpq2bt0qSdq5c6fuvPNO9e/fX3feeae++uqrwH3sGAMAAAAANE7ICmROTo5WrFihZcuWafjw4ZowYYIkKSsrS2lpacrPz1daWpoyMzMD97FjDAAAAADQOBGhCoqJiQn8f2VlpRwOh0pLS1VQUKBXXnlFkjRw4EBNnTpVZWVlsiwr6GNxcXGh+nTPSHGxkXJ5PLbN76uuVtm+w7bNDwAAAMBeISuQkjRx4kRt2LBBlmXp5ZdfVlFRkRISEuRyuSRJLpdLbdu2VVFRkSzLCvoYBfLEXB6PvssdY9v8HcbkSqJAAgAAAGeqkBbI6dOnS5KWLVummTNnauzYsaGMPyXx8S1CltWmTczJb3QWZIYzFwAAAMDpC2mBPGrw4MHKzMxUu3btVFxcLJ/PJ5fLJZ/Pp927dysxMVGWZQV97FSUllbK77dCUnhKSiqOO2Z3bjgyG8oFAAAA0HQ4nY4GT6iF5EV0Dhw4oKKiosDHa9euVWxsrOLj45WUlKRVq1ZJklatWqWkpCTFxcXZMgYAAAAAaLyQnIE8dOiQxo4dq0OHDsnpdCo2NlZ5eXlyOBx6+umnlZGRoRdeeEEtW7ZUTk5O4H52jAEAAAAAGickBbJ169Z6/fXX6x3r3Lmz/vKXv4RsDAAAAADQOCF7H0gAAAAAwJmNAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxEhHsBQKtYjyI8kbbNX1t9WHv3Vds2PwAAAHCuoEAi7CI8kdrywq22zd9t9ApJFEgAAADgdHEJKwAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjEeFeABAurWI9ivBE2jZ/bfVh7d1Xbdv8AAAAQKhRIHHOivBE6h8vDbRt/j4PrJJEgQQAAMDZIyQFcu/evXryySf1zTffyOPx6Pzzz1d2drbi4uLUtWtXdenSRU7nkatpZ86cqa5du0qS1q5dq5kzZ8rn8+mSSy7RjBkz1KxZs9MaAwAAAAA0TkieA+lwODRixAjl5+dr5cqVOu+88zR79uzA+KJFi7R8+XItX748UB4PHDigyZMnKy8vT2vWrFHz5s01f/780xoDAAAAADReSAqk1+tV7969Ax9ffvnl2rVr1wnv895776lbt27q1KmTJCk1NVVvvvnmaY0BAAAAABov5M+B9Pv9WrhwoVJSUgLH7r77bvl8Pl177bVKT0+Xx+NRUVGR2rdvH7hN+/btVVRUJEmNHgMAAAAANF7IC+TUqVMVHR2tu+66S5L0zjvvKDExUZWVlXriiSeUm5urcePGhXpZx4mPbxGyrDZtYkKWFc5McgEAAIAzW0gLZE5Ojr7++mvl5eUFXjQnMTFRktSiRQsNHTpUr7zySuD4xo0bA/fdtWtX4LaNHTsVpaWV8vutkBSAkpKK447ZnRuOTHIBAACAps/pdDR4Qi0kz4GUpDlz5mjLli3Kzc2Vx+ORJO3bt09VVVWSpNraWuXn5yspKUmS1KdPH23evFlfffWVpCMvtDNgwIDTGgMAAAAANF5IzkBu375deXl56tSpk1JTUyVJHTt21IgRI5SZmSmHw6Ha2lr16NFDY8eOlXTkjGR2drYefPBB+f1+JSUlaeLEiac1BgAAAABovJAUyIsuukjbtm2rd2zlypUN3u+GG27QDTfcENQxAAAAAEDjhOwSVgAAAADAmY0CCQAAAAAwQoEEAAAAABihQAIAAAAAjFAgAQAAAABGKJAAAAAAACMUSAAAAACAEQokAAAAAMAIBRIAAAAAYIQCCQAAAAAwQoEEAAAAABihQAIAAAAAjESEewHAucYb65HbE2nb/DXVh1W+r9q2+QEAAHDuokACIeb2RGr17wfYNv/Nw9+URIEEAABA8HEJKwAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBHjAnnNNdfUe/z6668/6X337t2rBx54QP3799ctt9yihx9+WGVlZZKkTz75RLfeeqv69++v4cOHq7S0NHA/O8YAAAAAAI1jXCAPHDhw3LHa2lodPHjwpPd1OBwaMWKE8vPztXLlSp133nmaPXu2LMvSE088oczMTOXn5ys5OVmzZ8+WJFvGAAAAAACNF3GyGwwbNkwOh0M1NTUaPnx4nbGioiJ17979pCFer1e9e/cOfHz55Zdr4cKF2rx5syIjI5WcnCxJSk1NVd++fTVjxgxbxoBzmdfrltsdZdv8NTVVKi+vsW1+AAAAhN9JC2T//v0lSZs2bVK/fv0Cxx0Oh+Lj43XttdeeUqDf79fChQuVkpKioqIitW/fPjAWFxcnv9+v8vJyW8a8Xu8prRU4m7jdUVr4an/b5v/lffmSKJAAAABns5MWyNTUVEnSZZddpqSkpNMOnDp1qqKjo3XXXXdpzZo1pz2fXeLjW4Qsq02bmJBlhTOTXHIBAABwZjtpgTwqKSlJH374obZu3Xrc8x4feughozlycnL09ddfKy8vT06nU4mJidq1a1dgvKysTA6HQ16v15axU1FaWim/3wrJL8QlJRXHHbM7NxyZ5J6buQAAADizOJ2OBk+oGb+IzowZMzR69Gj985//1Oeffx74b/v27Ub3nzNnjrZs2aLc3Fx5PB5JUrdu3VRVVaVNmzZJkhYtWqQBAwbYNgYAAAAAaDzjM5DLli3T0qVL1bFjx1MO2b59u/Ly8tSpU6fAJbEdO3ZUbm6uZs6cqaysLB0+fFgdOnTQrFmzJElOpzPoYwAAAACAxjMukC1btlRcXFyjQi666CJt27at3rGePXtq5cqVIRsDAAAAADSO8SWsI0eOVEZGhgoKClRcXFznPwAAAADA2c/4DOTkyZMlSW+//Xad4w6HQ1u3bg3uqgAAAAAATY5xgfz000/tXAcAAAAAoIkzvoS1oqJC1dXV8ng8gf+qq6tVWVlp5/oAAAAAAE2EcYF86KGH9O2339Y5VlhYqFGjRgV9UQAAAACApse4QO7cuVM/+clP6hxLSkrSF198EfRFAQAAAACaHuMC6fV6VVhYWOdYYWGhYmNjg74oAAAAAEDTY1wgBw8erLFjx+r9999XYWGhNmzYoEcffVS33XabnesDAAAAADQRxq/COmrUKDkcDk2aNEnFxcVKSEjQHXfcoZEjR9q5PgAAAABAE2FcIF0ul8aMGaMxY8Y0eJtXX31V9913XzDWBQAAAABoYowvYTUxb968YE4HAAAAAGhCjM9AmrAsK5jTATgLxHrd8rijbJu/uqZK+8prbJsfAAAA/yeoBdLhcARzOgBnAY87Si/+ob9t8z94d74kCiQAAEAoBPUSVgAAAADA2SuoBZJLWAEAAADg7BXUAtm9e/dgTgcAAAAAaEJO6TmQmzZt0urVq1VaWqp58+apoKBABw8eVHJysiTptddes2WRAAAAAIDwMz4DuWjRIj355JOKi4vThg0bJB15b8hnn33WtsUBAAAAAJoO4wI5f/58vfLKK0pPT5fTeeRunTt31o4dO2xbHAAAAACg6TAukBUVFerYsaOk/3u7Dr/fr4iIoL4TCAAAAACgiTIukD179tSrr75a59jChQt1xRVXBHtNAAAAAIAmyPj04eTJkzVy5Ei9/vrrOnjwoG699VZJ0ksvvWTb4gAAAAAATYdxgUxMTNTy5cv173//W999950SExPVq1cvLmEFAAAAgHOEUfvz+XxKTk7Wxo0bdcUVV3DZKgAAAACcg4yeA+lyudShQwft37/f7vUAAAAAAJoo4+tPhwwZolGjRmnYsGFKSEgIvBKrdOQFdgAAAAAAZzfjAvn73/9ekvTMM88cN7Z+/frgrQgAAAAA0CQZF0hKIgAAAACc24zfBxIAAAAAcG4zPgPZr1+/Os97PFZ+fn7QFgQAAAAAaJqMC+TkyZPrfFxSUqI//vGPuuWWW4K+KAAAAABA02NcIPv06XPcsauvvjrwyqwAAAAAgLPbaT0HMjo6Wt98802w1gIAAAAAaMKMz0Dm5eXV+fjQoUN65513dNVVVwV9UQAAAACApse4QH7++ed1Pm7WrJmGDh2q22+/PeiLAoDTFet1y+OOsm3+6poq7SuvsW1+AACApsi4QD777LN2rgMAgsrjjlLOov62zf+r1HxJFEgAAHBuMS6QkvTtt99q27ZtOnjwYJ3jvBIrAAAAAJz9jAvk/PnzNWfOHF144YVq1qxZ4LjD4aBAAgAAAMA54JQK5MKFC3XppZfauR4AAAAAQBNl/DYebrdbXbt2bXRQTk6OUlJS1LVr1zovyJOSkqKbbrpJgwYN0qBBg/SPf/wjMPbJJ5/o1ltvVf/+/TV8+HCVlpae9hgAAAAAoHGMC+TYsWOVk5Ojffv2NSqob9++WrBggTp06HDc2Lx587R8+XItX75cffr0kSRZlqUnnnhCmZmZys/PV3JysmbPnn1aYwAAAACAxjMukJMnT9aCBQt05ZVXqlu3burWrZsuueQSdevWzej+ycnJSkxMNF7Y5s2bFRkZqeTkZElSamqq3nrrrdMaAwAAAAA0nvFzIFeuXGnbIsaPHy/LstSrVy899thjatmypYqKitS+ffvAbeLi4uT3+1VeXt7oMa/Xa9vnAAAAAABnO+MCeeGFFwb+P5hlbMGCBUpMTFR1dbWmT5+u7OzsJnHJaXx8i5BltWkTE7KscGaSSy65AAAAZzbjAllZWanp06frzTfflHTkhWrWrVungoICjRkzptELOHpZq8fjUVpamkaNGhU4vmvXrsDtysrK5HA45PV6Gz12KkpLK+X3WyH5BbGkpOK4Y3bnhiOTXHLPhVwAAIAzndPpaPCEmvFzIKdMmSLpyKWsbrdbktS9e3etWLGi0Qs7ePCgKiqO/AJmWZbeeOMNJSUlSZK6deumqqoqbdq0SZK0aNEiDRgw4LTGAAAAAACNZ3wGcv369Xr33Xfl8XjkcDgkSfHx8dqzZ4/R/adNm6a3335be/bs0bBhw+T1epWXl6f09HT5fD75/X517txZWVlZkiSn06mZM2cqKytLhw8fVocOHTRr1qzTGgMAAAAANJ5xgWzevLn279+v1q1bB459//33dT4+kUmTJmnSpEnHHV+2bFmD9+nZs2eDL97T2DEAAAAAQOMYX8J622236dFHH9XHH38sy7JUUFCgCRMmaOjQoXauDwAAAADQRBifgRw1apTcbrfGjx+vgwcP6uGHH1Zqaqruv/9+O9cHAAAAAGgiTlogV61apYEDB8rpdGrkyJEaOXJkKNYFAAAAAGhiTnoJa2ZmZijWAQAAAABo4k5aIC3LCsU6AAAAAABN3EkvYfX7/frggw9OWCSvuuqqoC4KAAAAAND0nLRAVldXa+LEiQ0WSIfDob///e9BXxgAAAAAoGk5aYFs1qwZBREAAAAAYP4+kAAAAACAcxsvogMAAAAAMHLSAvnxxx+HYh0AAAAAgCbupM+BBACYa+n1KNIdadv8h2sOa395tW3zAwAAnAgFEgCCKNIdqbGLb7Jt/rm3vyWJAgkAAMKDF9EBAAAAABihQAIAAAAAjFAgAQAAAABGKJAAAAAAACMUSAAAAACAEQokAAAAAMAIBRIAAAAAYIQCCQAAAAAwQoEEAAAAABihQAIAAAAAjFAgAQAAAABGKJAAAAAAACMUSAAAAACAEQokAAAAAMAIBRIAAAAAYIQCCQAAAAAwQoEEAAAAABihQAIAAAAAjFAgAQAAAABGKJAAAAAAACMUSAAAAACAkYhwLwAAcPpivB5FuSNtm7+q5rAqyqttmx8AAJwZKJAAcBaIckdqwPJ7bZv/zUGvqUIUSAAAznVcwgoAAAAAMEKBBAAAAAAYoUACAAAAAIyEpEDm5OQoJSVFXbt21eeffx44vnPnTt15553q37+/7rzzTn311Ve2jgEAAAAAGi8kBbJv375asGCBOnToUOd4VlaW0tLSlJ+fr7S0NGVmZto6BgAAAABovJAUyOTkZCUmJtY5VlpaqoKCAg0cOFCSNHDgQBUUFKisrMyWMQAAAADA6Qnb23gUFRUpISFBLpdLkuRyudS2bVsVFRXJsqygj8XFxYXnEwUAAACAswTvA9mA+PgWIctq0yYmZFnhzCSXXHLJBQAAZ7awFcjExEQVFxfL5/PJ5XLJ5/Np9+7dSkxMlGVZQR87VaWllfL7rZD8wlRSUnHcMbtzw5FJLrnknn25AADg7ON0Oho8oRa2t/GIj49XUlKSVq1aJUlatWqVkpKSFBcXZ8sYAAAAAOD0hOQM5LRp0/T2229rz549GjZsmLxer1avXq2nn35aGRkZeuGFF9SyZUvl5OQE7mPHGAAAAACg8UJSICdNmqRJkyYdd7xz5876y1/+Uu997BgDAAAAADRe2C5hBQAAAACcWSiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxEhHsBAIAzV4w3UlFujy1zV9VUq6L8sC1zAwCAxqFAAgAaLcrt0S+WPm3L3G/c9rQqRIEEAKAp4RJWAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEYokAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEYokAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAAAAAEYiwr0AAABORYw3SlFut23zV9XUqKK8yrb5AQA4k1EgAQBnlCi3Wzcvec62+VcPeVQVokACAFAfLmEFAAAAABihQAIAAAAAjFAgAQAAAABGKJAAAAAAACNN4kV0UlJS5PF4FBkZKUkaP368+vTpo08++USZmZk6fPiwOnTooFmzZik+Pl6SGj0GAAAAAGicJnMGct68eVq+fLmWL1+uPn36yLIsPfHEE8rMzFR+fr6Sk5M1e/ZsSWr0GAAAAACg8ZpMgfyhzZs3KzIyUsnJyZKk1NRUvfXWW6c1BgAAAABovCZxCat05LJVy7LUq1cvPfbYYyoqKlL79u0D43FxcfL7/SovL2/0mNfrDennBAAAAABnkyZRIBcsWKDExERVV1dr+vTpys7O1o033hjWNcXHtwhZVps2MSHLCmcmueSSS+6ZkBnOXAAAmromUSATExMlSR6PR2lpaRo1apTuuece7dq1K3CbsrIyORwOeb1eJSYmNmrsVJSWVsrvt0LyS0RJScVxx+zODUcmueSSS+6ZkNlQLgAA5wqn09HgCbWwPwfy4MGDqqg48oPasiy98cYbSkpKUrdu3VRVVaVNmzZJkhYtWqQBAwZIUqPHAAAAAACNF/YzkKWlpUpPT5fP55Pf71fnzp2VlZUlp9OpmTNnKisrq87bcUhq9BgAAAAAoPHCXiDPO+88LVu2rN6xnj17auXKlUEdAwAAAAA0TtgvYQUAAAAAnBkokAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGAn7+0ACAHAmiPFGKcrttm3+qpoaVZRX2TY/AADBQIEEAMBAlNutmxe/bNv8q28foQpRIAEATRuXsAIAAAAAjFAgAQAAAABGKJAAAAAAACMUSAAAAACAEQokAAAAAMAIBRIAAAAAYIQCCQAAAAAwQoEEAAAAABihQAIAAAAAjFAgAQAAAABGIsK9AAAA0LAYb5Si3G7b5q+qqVFFeVU9uc0U5bbv14SqmlpVlB8i1xutKLfLxlyfKsoPHne8pTdakTbmHq7xaX89uV5vc7nd9py/qKnxq7z8QEgzT5TbyttcETbm1tb4tbe+3NjmivDYmFvt1959x+fi3EGBBACgCYtyuzXwrwtsm3/VHf+/KnR8gYxyR+iWvy61LXflHbepop7jUe4IDfprvm25y+/o32DukMXv25a75PafNZDr0v+3+H9sy3399p/Umxvpdilr6S7bcqfc1r7e4263U68tKbEl894hbRrMfPPPe2zJlKQBd7au93iE26kPXt1tW+6V97WtP9fj1Oe5xbbldhmTYNvcODNQIAEAAACclrjYaLk89p3V9lX7VLbv+LPaCD0KJAAAAIDT4vK49P2sr22bv90T59d7nOIaehRIAAAAAGckl8el4jmf2DZ/wrjLbZv7TEWBBAAAAIBTEBfbTC6PfVXKV12rsn3Hv/BWU0CBBAAAAIBT4PJEqHjee7bNn/DItbbNfbp4H0gAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMEKBBAAAAAAYoUACAAAAAIxQIAEAAAAARiiQAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMBIR7gXYZefOncrIyFB5ebm8Xq9ycnLUqVOncC8LAAAAABolLraZXB77KpyvulZl+w6d8DZnbYHMyspSWlqaBg0apOXLlyszM1P//d//He5lAQAAAECjuDwR2v38Ktvmb/vwwJPe5qy8hLW0tFQFBQUaOPDIF2DgwIEqKChQWVlZmFcGAAAAAGeus/IMZFFRkRISEuRyuSRJLpdLbdu2VVFRkeLi4ozmcDod//f/Mc1tWWd9WXWOx7QMeaYrxuzrE+xcd0zbsORGtghPbrMw5TZvkRCW3BbNw5PbMky5cdHhyW3brHV4cqO9Yci0b388cW6LMOWG5+dQ2+joMOVGhSW3TXRkmHLdYcn1RrvCkts82r7zFw3+3LUx80S5kS3CkxsRE55cV8vwPKacLT3hyY0Jz57hjGlme25D2ZLksCzLsnUFYbBlyxb96le/0urVqwPHfvGLX2jWrFm65JJLwrgyAAAAADhznZWXsCYmJqq4uFg+n0+S5PP5tHv3biUmJoZ5ZQAAAABw5jorC2R8fLySkpK0atWRJ5iuWrVKSUlJxpevAgAAAACOd1ZewipJO3bsUEZGhvbv36+WLVsqJydHF154YbiXBQAAAABnrLO2QAIAAAAAguusvIQVAAAAABB8FEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMBIR7gWcqfbu3asnn3xS33zzjTwej84//3xlZ2frq6++0pQpUwK3Ky0tVZs2bbR06VLbMuPi4vTXv/5Vr732mpxOp1wulyZMmKDk5OTTzjxZbnl5ubKzs/XZZ58pIiJCAwYM0MMPP2x7bteuXdWlSxc5nUf+BjJz5kx17drV9tyjnnrqKS1ZskQfffSRmjdvbmuu1+vVL3/5Sx06dEiS1KZNG02ZMkUdO3a0NXffvn3KzMxUSUmJIiIidOmllyorK0tRUVG25sbFxenxxx/Xxo0bVVJSEtSv8clyP/nkE2VmZurw4cPq0KGDZs2apfj4+KDk5uTkKD8/X999951WrlypLl26SJLeeecdzZ07V7W1tYqNjdWMGTN03nnnBSXzRLnr1q3T3LlzZVmW/H6/0tPT1a9fP1tzv/32W40ZMyZwm4qKClVWVupf//qXrbmSdPjwYT3zzDP65z//qcjISF1++eWaOnWq7bkpKSnyeDyKjIyUJI0fP159+vQJWm59nn/+ef3mN7+psw47jR49Wt9++62cTqeio6M1efJkJSUl2Z5r92P4qPr+bU1+TgSb3XtFQ0L1GG7oe2jnzp3KyMhQeXm5vF6vcnJy1KlTJ9tzGzpuF7v3qLUr47MAABCcSURBVPqEYk+WGv5ahnLvCOX37MkeO6HYo21//FpolL1791offPBB4ONf//rX1lNPPXXc7UaNGmW9/PLLtmaWlZVZPXr0sEpKSizLsqy//e1v1oABA4KSeaJcy7KsBx980HrllVcCY0fXYHduly5drMrKyqBlmeZalmX9/e9/t5566qmgr+FEufv37w8cf/XVV60xY8bYnltYWGh99tlnlmVZls/ns8aOHWs9//zztudalmW9//771p49e2z5d24o1+/3WzfccIP14YcfWpZlWbm5uVZGRkbQcj/88ENr165d1s9//nNr27ZtlmVZVnl5ufXTn/7U+vLLLy3Lsqxly5ZZw4cPD1pmQ7l+v99KTk4OfLx161br8ssvt3w+n625PzRt2jRrypQpQcs8Ue7UqVOt6dOnW36/37Ks4O5VJ8o90edvhy1btlj333+/df3114cs99j9ac2aNdbgwYNtzwzFY/io+v5tTX8HCJZQ7BUNCdVjuKHvobvvvttatmyZZVlHPu+77747JLkme1gw2b1HmbBjT7ashr+Wodw7Qvk9e6LHTqj2aLsfv1zC2kher1e9e/cOfHz55Zdr165ddW5TWlqqDRs2aNCgQbZmWv/7Vp4HDhyQdOQvSO3atQtK5olyv/rqK33++ee69957A2OtW7e2PdduJ8rdu3evnn/+eT311FMhzY2JiQkcr6ysDJx5tTO3Y8eOuvjiiyVJTqdT3bt3D+rX/0Sf71VXXRW0M3+muZs3b1ZkZGTgzH1qaqreeuutoOUmJycrMTGxzrGvv/5arVu31gUXXCBJuu6667R+/XqVlZXZmisd+TetqKiQdGTPaNu2bVAfVw3lHlVdXa2VK1fq9ttvD1pmQ7kHDhzQsmXLNHbsWDkcDknB3asayg216upqZWdnKysrK/B5hsIP96dQZdv9GD6qvn/bUP98CsVeEW71fZ1LS0tVUFCggQMHSpIGDhyogoKCkOyRofyeDsUedTJ27clSw1/LUO4dofyebejzDeUebffjl0tYg8Dv92vhwoVKSUmpc3zZsmW6+uqrbdkEjs2Mi4vT008/rcGDBys2NlZ+v19/+MMfgp75w9wvvvhCCQkJmjhxorZu3arWrVvrySef1EUXXWRr7lF33323fD6frr32WqWnp8vj8diem52drfT09Dqbnh3q+3wfeOABFRQUqFWrVpo/f37IciWpqqpKixcv1mOPPRbSXLsdm1tUVKT27dsHxuLi4uT3+wOXTdnhggsu0J49e/Tpp5+qe/fuWrlypSSpqKjI1kvhHA6HnnvuOY0ePVrR0dE6cOCAXnzxRdvy6rN27VolJCTokksusT2rsLBQXq9Xzz//vDZu3KjmzZtr7NixQbvM/2TGjx8vy7LUq1cvPfbYY2rZsqUtOXPnztWtt94akssaf2jixInasGGDLMvSyy+/bHteU3gMHxWK/Stce8VRoXoM/1BRUZESEhLkcrkkSS6XS23btg3Z5x0q4d6jpNDuyccK9d4hhe93jnDu0cHGGcggmDp1qqKjo3XXXXfVOb5kyRJb/pLzw8zKykr96U9/0uLFi/XOO+8oIyNDDz/8cODMpF25Pp9P//nPfzRkyBAtXbpUQ4cO1ahRo4Ke+cNc6chzQZYsWaIFCxboiy++UG5uru25b775ptxut37+85/bktVQ7lEvvfSS/vGPf+jmm2/Wb3/725Dl1tbWaty4cbryyivVt2/fkOWGQrhyj4qJidGcOXM0Y8YMDRkyRKWlpWrZsqUiIuz9215tba1efPFFvfDCC1q3bp1++9vfaty4cYGrGEJh8eLFtu2PP1RbW6vCwkJdfPHFWrJkicaPH6/09HRVVlbanr1gwQKtWLFCixcvlmVZys7OtiXn448/1ubNm5WWlmbL/Cczffp0vfPOOxo3bpxmzpxpe15TeAwfFYp9JFx7hRS6x/C5LJx71FGh3JOPFeq9QwrPz/5w79HBRoE8TTk5Ofr666/13HPP1bl05pNPPlF5ebmuu+462zPXr1+vmJgYXXjhhZKkX/ziF/rmm2+0d+9eW3Pbt2+vxMTEwF/I+vXrp5KSkqBfUlPf1/joafkWLVpo6NCh+uijj4KaWV/uxo0b9cEHHyglJSXwV6uBAwfqiy++sDX3WE6nU3fccYeWL18e1MyGcn0+n8aPH6/Y2FhNmjQp6JkN5YbCD3MTExPrXM5SVlYmh8Nh29nHo372s59p4cKFWrJkie666y5VVVXZ/tfJrVu3avfu3erVq5ckqVevXmrWrJl27Nhha+5RxcXF+vDDD3XLLbeEJK99+/aKiIgIXAZ32WWXqVWrVtq5c6ft2Uf3Ko/Ho7S0NFv2Kkn68MMP9eWXX6pv375KSUnR999/r/vvv1/r16+3Ja8hgwcP1saNG4P+8+eHwv0YPiqU+1c49gopdI/hhrKLi4vl8/kkHfmZtHv37rBfLh5s4dyjpNDvyfUJ1d4Rrt85msoeHSwUyNMwZ84cbdmyRbm5ucddPrl48WINGjQo6H8drC+zY8eO2rp1q0pLSyVJH3zwgVq0aKFWrVrZmtutWzdFR0dr+/btko58c8TGxtqeu2/fPlVVVUk68le7/Pz8oL9qV325Tz/9tN577z2tXbtWa9eulSStWrVKP/7xj23NLSsrq7OhvvXWW0F7xdkT5fr9fmVkZMjlcmn69Om2XK9/ou8hOzX0eK6qqtKmTZskSYsWLdKAAQNsX0tJSYmkI1/vZ599VqmpqYqOjrY1s127dvr+++/15ZdfSpJ27NihPXv26Ec/+pGtuUctXbpU1113XVD3ihOJi4tT7969tWHDBklHXtWxtLRU559/vq25Bw8eDDxHz7IsvfHGG7a9wuDIkSO1fv36wP7Url07zZ8/X9dcc40teUcdOHBARUVFgY/Xrl2r2NhY2//wEu7HsBT6/Ssce0UoH8P1iY+PV1JSklatWiXpyM/cpKSks+ryVSl8e9RRod6TpfDsHeH6nUMK3x5tF4dlx3WO54Dt27dr4MCB6tSpU+BtDTp27Kjc3FxVVVXpmmuu0Z///Gd17tw5JJmvvPKKXn/9dbndbnk8HmVkZATt2vkT5W7evFlTpkxRdXW1mjVrpokTJ6p79+625o4YMUKZmZlyOByqra1Vjx49NGHChKC91cOJPt9jde3aNahvMdFQ7iOPPKKnnnpKNTU1kqQOHTpo4sSJQfvLc0O5Q4cO1YMPPljn7VJ69uyprKwsW3Nzc3P18MMP69NPP1VxcbHatm2rLl26BO15nyfK/eijj5SVlVXnbTyC9RzmadOm6e2339aePXvUqlUreb1erV69WhMnTtRHH32kmpoaXX311ZowYULg5fLtzF2xYoVeeumlwB8GHnnkEd1www2250pS//79NXHiRF177bVByztZbmFhoSZMmKDy8nJFRETo0UcfDeoVIvXl5uXlKT09XT6fT36/X507d9akSZPUtm3boOU2JCUlRXl5eba/9cCePXs0evRoHTp0SE6nU7GxsfrVr34VkudR2f0YPqq+f9vnnnvO6OdEMNm9V9SnsLAwZI/hhr53d+zYoYyMDO3fv18tW7ZUTk5O4IorO3NPtIfZwe496kTs3JOl+r/Gr732Wkj3DtPf7YLB5LFj9x5t9+OXAgkAAAAAMMIlrAAAAAAAIxRIAAAAAIARCiQAAAAAwAgFEgAAAABghAIJAAAAADBCgQQAAAAAGKFAAgAAAACMUCABAOeUHj16BP77yU9+ou7duwc+XrFiRbiXd1Jr167VNddco/379weOVVVV6cYbb9TSpUvDuDIAwLnAYVmWFe5FAAAQDikpKZo2bZp+9rOfhTy7trZWERERjbrv448/rsjISD3zzDOSpJycHG3fvl0vv/xyMJd4WmsEAJydOAMJAMAxfD6fcnNz1bdvX/Xu3VuPP/544Gzfjh07dPHFF2vx4sW69tprdeWVV9YpbePGjdMLL7wQ+Pi9997TjTfeGPj46quv1vz583XzzTerV69ekqSioiKNGjVKvXv3Vt++fbVw4cKTrnHSpEl699139f777+uzzz7T4sWLlZ2dHRj/17/+pTvuuEPJyckaMmSIPv7448DYggUL1L9/f/Xo0UP9+vXTsmXLAmPr1q3TTTfdpHnz5umqq67S1KlTVVxcrPvvv1/Jycnq3bu3hg8f3oivKgDgbMGfFQEAOMb8+fO1YcMG/elPf5LX61VWVpZmzJihGTNmSDpSMLds2aI1a9Zo27ZtSktLU//+/XXeeecZzf/GG2/o97//vWJjY+Xz+TRy5Ejdcsstmjt3rr777jsNGzZMnTt31k9/+tMG52jVqpUmTZqkyZMnKyYmRuPGjVP79u0lSd98843S09P13HPPqXfv3lq3bp1Gjx6tt99+WzExMUpISND8+fPVoUMHrV+/XmPGjNGll16qzp07B+7vdDr17rvvyu/3a9asWbrooov04osvyu/36z//+c9pfoUBAGcyzkACAHCMRYsW6fHHH1dCQoIiIyM1ZswYvfHGGzr2GR/p6emKjIxU9+7ddcEFF2jbtm3G8993331KSEhQVFSU/v3vf+vw4cMaOXKkPB6PLrjgAg0ZMkSrV68+6TwDBgzQRRddpKioKKWmpgaOL1myRDfddJOuuuoqOZ1O9e3bV+eff77ef/99SdINN9ygjh07yuFwqE+fPurZs6c++uijwP2joqL00EMPyePxKCoqShERESouLlZRUZE8Ho+uuOIK488VAHD24QwkAAD/y7Isff/99xo5cqQcDkfguN/v1969eyVJLpdLcXFxgbFmzZrpwIEDxhmJiYmB/9+1a5e+++47JScnB475fD7j52T++Mc/1t69e+usddeuXXrrrbfqlNDa2lrt3r1bkrRmzRrl5eWpsLBQfr9fVVVVdfLatGlT53mPo0eP1rx583T33XcrMjJSv/zlL3XfffcZf74AgLMLBRIAgP/lcDiUkJCg3/zmN+rWrdtx40dLZEOaNWumqqqqwMd79uw54e3btWunCy+8UCtXrmzcguuRmJio1NRUTZgw4bixyspKjRs3Trm5ubr66qsVERFxXBk8toxKUmxsrCZPnqzJkyeroKBA99xzj3r06KHLLrssaGsGAJw5uIQVAIBjpKam6r/+679UVFQkSSotLdXatWuN7puUlKR169Zp//79Ki4u1h//+McT3v7oC+m8+uqrOnz4sGpra/U///M/+uyzzxq9/ttuu02rV6/WBx98EDjD+P7772vPnj2qqqpSbW2t4uLi5HQ6tWbNGm3atOmE8/3tb39TYWGhLMtSTEyMnE6nnE5+fQCAcxVnIAEAOMaIESPkcDh07733qqSkRK1bt9agQYOUkpJy0vvefvvt2rhxo66//nr96Ec/0i233KJFixY1eHu3263f/e53+vWvf63f/e53qqmpUefOnfX44483ev2dOnXS3LlzNXv2bO3YsUNut1uXXXaZsrOz1aZNG40fP14PPPCAfD6f+vXrpz59+pxwvi+++ELTpk3Tvn375PV69eCDD+rSSy9t9PoAAGc23gcSAAAAAGCEa1AAAAAAAEa4hBUAgCbonnvu0ebNm487/sgjj2jYsGFhWBEAAFzCCgAAAAAwxCWsAAAAAAAjFEgAAAAAgBEKJAAAAADACAUSAAAAAGCEAgkAAAAAMPL/ADE0l3r8zKXQAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (15, 6)})\n",
    "sns.barplot(data=q11, x='Tenure_Years', y='Tenure_cnt',\\\n",
    "           order=q11.sort_values('Tenure_cnt',ascending = False).Tenure_Years)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **12. Count of Employee Status (Currently working or Left) in different departments grouped by gender**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------------------+---+-----------+-------------+----------+\n",
      "|         dept_name|sex|Total_Count|Working_Count|Left_Count|\n",
      "+------------------+---+-----------+-------------+----------+\n",
      "|Quality Management|  M|      12039|        10839|      1200|\n",
      "|           Finance|  F|       5961|         5388|       573|\n",
      "|   Human Resources|  M|      10711|         9635|      1076|\n",
      "|Quality Management|  F|       8078|         7260|       818|\n",
      "|   Human Resources|  F|       7075|         6354|       721|\n",
      "|             Sales|  M|      29810|        26755|      3055|\n",
      "|  Customer Service|  M|      14132|        12713|      1419|\n",
      "|       development|  M|      47277|        42512|      4765|\n",
      "|         Marketing|  F|       5850|         5304|       546|\n",
      "|          Research|  F|       8439|         7601|       838|\n",
      "|        Production|  F|      23912|        21460|      2452|\n",
      "|          Research|  M|      12687|        11427|      1260|\n",
      "|           Finance|  M|       8756|         7929|       827|\n",
      "|             Sales|  F|      19851|        17954|      1897|\n",
      "|  Customer Service|  F|       9448|         8453|       995|\n",
      "|       development|  F|      31437|        28370|      3067|\n",
      "|        Production|  M|      35577|        32069|      3508|\n",
      "|         Marketing|  M|       8984|         8134|       850|\n",
      "+------------------+---+-----------+-------------+----------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "            SELECT\n",
    "                dept_name, sex,\n",
    "                count(Left2) Total_Count,\n",
    "                sum(CASE WHEN Left2 = 0 THEN 1 ELSE 0 END) Working_Count,\n",
    "                sum(Left2) Left_Count\n",
    "            FROM Employees e\n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no\n",
    "            JOIN departments d ON de.dept_no = d.dept_no\n",
    "            GROUP BY dept_name, sex\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 13. Max, Min and Avg age of Employees in diffrent departments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------------------+--------+--------+------------------+\n",
      "|         dept_name|min(Age)|max(Age)|          avg(Age)|\n",
      "+------------------+--------+--------+------------------+\n",
      "|             Sales|      48|      61| 54.91617164374459|\n",
      "|Quality Management|      48|      61| 54.94397773027787|\n",
      "|           Finance|      48|      61|  54.9266154786981|\n",
      "|        Production|      48|      61| 54.93388693708081|\n",
      "|          Research|      48|      61|54.902821168228726|\n",
      "|  Customer Service|      48|      61| 54.91675148430873|\n",
      "|         Marketing|      48|      61| 54.92153161655656|\n",
      "|       development|      48|      61|  54.8962446324669|\n",
      "|   Human Resources|      48|      61| 54.94900483526369|\n",
      "+------------------+--------+--------+------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "spark.sql(\"\"\"\n",
    "           SELECT dept_name, min(Age), max(Age), avg(Age)\n",
    "           FROM (\n",
    "            SELECT\n",
    "                dept_name, e.emp_no,\n",
    "                    (SELECT year(greatest(max(last_date), max(hire_date))) FROM Employees)-year(birth_date) Age\n",
    "            FROM employees e\n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no\n",
    "            JOIN departments d ON de.dept_no = d.dept_no \n",
    "            )a\n",
    "          GROUP BY dept_name\n",
    "         \"\"\").show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Age group distribution is very uniform accross all the departments. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14. Count of Employees in various titles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>title</th>\n",
       "      <th>Emp_Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Staff</td>\n",
       "      <td>107384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>97747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Engineer</td>\n",
       "      <td>47303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Senior Staff</td>\n",
       "      <td>26583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Technique Leader</td>\n",
       "      <td>15148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Assistant Engineer</td>\n",
       "      <td>5835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Manager</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                title  Emp_Count\n",
       "0               Staff     107384\n",
       "1     Senior Engineer      97747\n",
       "2            Engineer      47303\n",
       "3        Senior Staff      26583\n",
       "4    Technique Leader      15148\n",
       "5  Assistant Engineer       5835\n",
       "6             Manager         24"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q14 = spark.sql(\"\"\"\n",
    "            SELECT \n",
    "                t.title, \n",
    "                count(Salary) Emp_Count\n",
    "            FROM salaries s\n",
    "            JOIN employees e ON s.emp_no = e.emp_no\n",
    "            JOIN titles t ON e.emp_title_id = t.title_id\n",
    "            GROUP BY t.title\n",
    "            ORDER BY Emp_Count DESC\n",
    "         \"\"\").toPandas()\n",
    "q14.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (15, 6)})\n",
    "sns.barplot(data=q14, x='title', y='Emp_Count')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 15 Average Tenure Distribution accross Departments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>dept_name</th>\n",
       "      <th>Tenure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Finance</td>\n",
       "      <td>21.911123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Marketing</td>\n",
       "      <td>21.899285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Research</td>\n",
       "      <td>21.864480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Production</td>\n",
       "      <td>21.835213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sales</td>\n",
       "      <td>21.832484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>development</td>\n",
       "      <td>21.829598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Quality Management</td>\n",
       "      <td>21.800517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Customer Service</td>\n",
       "      <td>21.790967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Human Resources</td>\n",
       "      <td>21.787473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            dept_name     Tenure\n",
       "0             Finance  21.911123\n",
       "1           Marketing  21.899285\n",
       "2            Research  21.864480\n",
       "3          Production  21.835213\n",
       "4               Sales  21.832484\n",
       "5         development  21.829598\n",
       "6  Quality Management  21.800517\n",
       "7    Customer Service  21.790967\n",
       "8     Human Resources  21.787473"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q15 = spark.sql(\"\"\"\n",
    "            SELECT \n",
    "                dept_name, \n",
    "                avg(Tenure_Years) Tenure\n",
    "            FROM employees_at e\n",
    "            JOIN dept_emp de ON e.emp_no = de.emp_no\n",
    "            JOIN departments d ON de.dept_no = d.dept_no\n",
    "            GROUP BY dept_name\n",
    "            ORDER BY Tenure DESC\n",
    "         \"\"\").toPandas()\n",
    "q15.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (15, 6)})\n",
    "sns.barplot(data=q15, x='dept_name', y='Tenure')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 16 Average Tenure Distribution accross Titles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>title</th>\n",
       "      <th>Tenure</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Manager</td>\n",
       "      <td>24.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>22.490163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Senior Staff</td>\n",
       "      <td>21.867961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Staff</td>\n",
       "      <td>21.830664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Technique Leader</td>\n",
       "      <td>21.810536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Engineer</td>\n",
       "      <td>20.638416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Assistant Engineer</td>\n",
       "      <td>20.470094</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                title     Tenure\n",
       "0             Manager  24.666667\n",
       "1     Senior Engineer  22.490163\n",
       "2        Senior Staff  21.867961\n",
       "3               Staff  21.830664\n",
       "4    Technique Leader  21.810536\n",
       "5            Engineer  20.638416\n",
       "6  Assistant Engineer  20.470094"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "q16 = spark.sql(\"\"\"\n",
    "            SELECT \n",
    "                title, \n",
    "                avg(Tenure_Years) Tenure\n",
    "            FROM employees_at e\n",
    "            JOIN titles t ON e.emp_title_id = t.title_id\n",
    "            GROUP BY title\n",
    "            ORDER BY Tenure DESC\n",
    "         \"\"\").toPandas()\n",
    "q16.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (15, 6)})\n",
    "sns.barplot(data=q16, x='title', y='Tenure')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Building Spark ML Model and Pipeline**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Creating Final Dataframe**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>dept_no</th>\n",
       "      <th>emp_no</th>\n",
       "      <th>emp_title_id</th>\n",
       "      <th>birth_date</th>\n",
       "      <th>first_name</th>\n",
       "      <th>last_name</th>\n",
       "      <th>sex</th>\n",
       "      <th>hire_date</th>\n",
       "      <th>no_of_projects</th>\n",
       "      <th>last_performance_rating</th>\n",
       "      <th>left</th>\n",
       "      <th>last_date</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure_Years</th>\n",
       "      <th>title_id</th>\n",
       "      <th>title</th>\n",
       "      <th>salary</th>\n",
       "      <th>dept_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>d005</td>\n",
       "      <td>10001</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-09-02</td>\n",
       "      <td>Georgi</td>\n",
       "      <td>Facello</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-06-26</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "      <td>1994-07-30</td>\n",
       "      <td>60</td>\n",
       "      <td>8</td>\n",
       "      <td>e0003</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>60117</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>d007</td>\n",
       "      <td>10002</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1964-06-02</td>\n",
       "      <td>Bezalel</td>\n",
       "      <td>Simmel</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-11-21</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>49</td>\n",
       "      <td>28</td>\n",
       "      <td>s0001</td>\n",
       "      <td>Staff</td>\n",
       "      <td>65828</td>\n",
       "      <td>Sales</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>d004</td>\n",
       "      <td>10003</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1959-12-03</td>\n",
       "      <td>Parto</td>\n",
       "      <td>Bamford</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-08-28</td>\n",
       "      <td>1</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>54</td>\n",
       "      <td>27</td>\n",
       "      <td>e0003</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40006</td>\n",
       "      <td>Production</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>d004</td>\n",
       "      <td>10004</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1954-05-01</td>\n",
       "      <td>Chirstian</td>\n",
       "      <td>Koblick</td>\n",
       "      <td>M</td>\n",
       "      <td>1986-12-01</td>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>59</td>\n",
       "      <td>27</td>\n",
       "      <td>e0003</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40054</td>\n",
       "      <td>Production</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>d003</td>\n",
       "      <td>10005</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1955-01-21</td>\n",
       "      <td>Kyoichi</td>\n",
       "      <td>Maliniak</td>\n",
       "      <td>M</td>\n",
       "      <td>1989-09-12</td>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>58</td>\n",
       "      <td>24</td>\n",
       "      <td>s0001</td>\n",
       "      <td>Staff</td>\n",
       "      <td>78228</td>\n",
       "      <td>Human Resources</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>d005</td>\n",
       "      <td>10006</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1953-04-20</td>\n",
       "      <td>Anneke</td>\n",
       "      <td>Preusig</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-06-02</td>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>60</td>\n",
       "      <td>24</td>\n",
       "      <td>e0003</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40000</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>d008</td>\n",
       "      <td>10007</td>\n",
       "      <td>s0001</td>\n",
       "      <td>1957-05-23</td>\n",
       "      <td>Tzvetan</td>\n",
       "      <td>Zielinski</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-02-10</td>\n",
       "      <td>6</td>\n",
       "      <td>B</td>\n",
       "      <td>1</td>\n",
       "      <td>2002-09-18</td>\n",
       "      <td>56</td>\n",
       "      <td>13</td>\n",
       "      <td>s0001</td>\n",
       "      <td>Staff</td>\n",
       "      <td>56724</td>\n",
       "      <td>Research</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>d005</td>\n",
       "      <td>10008</td>\n",
       "      <td>e0001</td>\n",
       "      <td>1958-02-19</td>\n",
       "      <td>Saniya</td>\n",
       "      <td>Kalloufi</td>\n",
       "      <td>M</td>\n",
       "      <td>1994-09-15</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>55</td>\n",
       "      <td>19</td>\n",
       "      <td>e0001</td>\n",
       "      <td>Assistant Engineer</td>\n",
       "      <td>46671</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>d006</td>\n",
       "      <td>10009</td>\n",
       "      <td>e0003</td>\n",
       "      <td>1952-04-19</td>\n",
       "      <td>Sumant</td>\n",
       "      <td>Peac</td>\n",
       "      <td>F</td>\n",
       "      <td>1985-02-18</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>61</td>\n",
       "      <td>28</td>\n",
       "      <td>e0003</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>60929</td>\n",
       "      <td>Quality Management</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>d006</td>\n",
       "      <td>10010</td>\n",
       "      <td>e0002</td>\n",
       "      <td>1963-06-01</td>\n",
       "      <td>Duangkaew</td>\n",
       "      <td>Piveteau</td>\n",
       "      <td>F</td>\n",
       "      <td>1989-08-24</td>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>50</td>\n",
       "      <td>24</td>\n",
       "      <td>e0002</td>\n",
       "      <td>Engineer</td>\n",
       "      <td>72488</td>\n",
       "      <td>Quality Management</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  dept_no  emp_no emp_title_id  birth_date first_name  last_name sex  \\\n",
       "0    d005   10001        e0003  1953-09-02     Georgi    Facello   M   \n",
       "1    d007   10002        s0001  1964-06-02    Bezalel     Simmel   F   \n",
       "2    d004   10003        e0003  1959-12-03      Parto    Bamford   M   \n",
       "3    d004   10004        e0003  1954-05-01  Chirstian    Koblick   M   \n",
       "4    d003   10005        s0001  1955-01-21    Kyoichi   Maliniak   M   \n",
       "5    d005   10006        e0003  1953-04-20     Anneke    Preusig   F   \n",
       "6    d008   10007        s0001  1957-05-23    Tzvetan  Zielinski   F   \n",
       "7    d005   10008        e0001  1958-02-19     Saniya   Kalloufi   M   \n",
       "8    d006   10009        e0003  1952-04-19     Sumant       Peac   F   \n",
       "9    d006   10010        e0002  1963-06-01  Duangkaew   Piveteau   F   \n",
       "\n",
       "    hire_date  no_of_projects last_performance_rating  left   last_date  Age  \\\n",
       "0  1986-06-26               9                       C     1  1994-07-30   60   \n",
       "1  1985-11-21               8                       B     0        None   49   \n",
       "2  1986-08-28               1                       C     0        None   54   \n",
       "3  1986-12-01               5                       A     0        None   59   \n",
       "4  1989-09-12               6                       A     0        None   58   \n",
       "5  1989-06-02              10                       B     0        None   60   \n",
       "6  1989-02-10               6                       B     1  2002-09-18   56   \n",
       "7  1994-09-15               9                       C     0        None   55   \n",
       "8  1985-02-18               8                       B     0        None   61   \n",
       "9  1989-08-24               4                       A     0        None   50   \n",
       "\n",
       "   Tenure_Years title_id               title  salary           dept_name  \n",
       "0             8    e0003     Senior Engineer   60117         development  \n",
       "1            28    s0001               Staff   65828               Sales  \n",
       "2            27    e0003     Senior Engineer   40006          Production  \n",
       "3            27    e0003     Senior Engineer   40054          Production  \n",
       "4            24    s0001               Staff   78228     Human Resources  \n",
       "5            24    e0003     Senior Engineer   40000         development  \n",
       "6            13    s0001               Staff   56724            Research  \n",
       "7            19    e0001  Assistant Engineer   46671         development  \n",
       "8            28    e0003     Senior Engineer   60929  Quality Management  \n",
       "9            24    e0002            Engineer   72488  Quality Management  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Joining employees, salaries, departments, titles\n",
    "emp_tsd = employees.join(titles, employees.emp_title_id == titles.title_id)\\\n",
    "    .join(salaries, on='emp_no')\\\n",
    "    .join(dept_emp, on='emp_no')\\\n",
    "    .join(departments, on='dept_no')\\\n",
    "    .withColumnRenamed('left2', 'left')\n",
    "emp_tsd.toPandas().head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "300024"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emp_tsd.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dept_no                    object\n",
       "emp_no                      int32\n",
       "emp_title_id               object\n",
       "birth_date                 object\n",
       "first_name                 object\n",
       "last_name                  object\n",
       "sex                        object\n",
       "hire_date                  object\n",
       "no_of_projects              int32\n",
       "last_performance_rating    object\n",
       "left                        int32\n",
       "last_date                  object\n",
       "Age                         int32\n",
       "Tenure_Years                int32\n",
       "title_id                   object\n",
       "title                      object\n",
       "salary                      int32\n",
       "dept_name                  object\n",
       "dtype: object"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfp = emp_tsd.toPandas()\n",
    "dfp.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "conti_var_df = dfp.loc[:, (dfp.dtypes == 'float64') | (dfp.dtypes == 'int64') | (dfp.dtypes == 'int32')]\n",
    "cat_var_df = dfp.loc[:, (dfp.dtypes == 'object')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Continous Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "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>dtype</th>\n",
       "      <th>tot</th>\n",
       "      <th>n</th>\n",
       "      <th>nmiss</th>\n",
       "      <th>miss_perc</th>\n",
       "      <th>cardinality</th>\n",
       "      <th>sum</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>var</th>\n",
       "      <th>iqr</th>\n",
       "      <th>lc_iqr</th>\n",
       "      <th>uc_iqr</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>p1</th>\n",
       "      <th>p5</th>\n",
       "      <th>p10</th>\n",
       "      <th>p25</th>\n",
       "      <th>p50</th>\n",
       "      <th>p75</th>\n",
       "      <th>p90</th>\n",
       "      <th>p95</th>\n",
       "      <th>p99</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>emp_no</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>300024</td>\n",
       "      <td>76002608740</td>\n",
       "      <td>253322</td>\n",
       "      <td>161828</td>\n",
       "      <td>2.61884e+10</td>\n",
       "      <td>339986</td>\n",
       "      <td>-424973</td>\n",
       "      <td>934973</td>\n",
       "      <td>10001</td>\n",
       "      <td>499999</td>\n",
       "      <td>13001.2</td>\n",
       "      <td>25002.2</td>\n",
       "      <td>40003.3</td>\n",
       "      <td>85006.8</td>\n",
       "      <td>249988</td>\n",
       "      <td>424993</td>\n",
       "      <td>469997</td>\n",
       "      <td>484998</td>\n",
       "      <td>496999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>no_of_projects</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>1651580</td>\n",
       "      <td>5.50483</td>\n",
       "      <td>2.87207</td>\n",
       "      <td>8.24881</td>\n",
       "      <td>5</td>\n",
       "      <td>-4.5</td>\n",
       "      <td>15.5</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>left</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>29867</td>\n",
       "      <td>0.0995487</td>\n",
       "      <td>0.299398</td>\n",
       "      <td>0.0896391</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>16476763</td>\n",
       "      <td>54.9181</td>\n",
       "      <td>3.74869</td>\n",
       "      <td>14.0527</td>\n",
       "      <td>6</td>\n",
       "      <td>43</td>\n",
       "      <td>67</td>\n",
       "      <td>48</td>\n",
       "      <td>61</td>\n",
       "      <td>49</td>\n",
       "      <td>49</td>\n",
       "      <td>50</td>\n",
       "      <td>52</td>\n",
       "      <td>55</td>\n",
       "      <td>58</td>\n",
       "      <td>60</td>\n",
       "      <td>61</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Tenure_Years</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>6550606</td>\n",
       "      <td>21.8336</td>\n",
       "      <td>5.92815</td>\n",
       "      <td>35.1429</td>\n",
       "      <td>6</td>\n",
       "      <td>11</td>\n",
       "      <td>35</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>14</td>\n",
       "      <td>20</td>\n",
       "      <td>23</td>\n",
       "      <td>26</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>salary</th>\n",
       "      <td>int32</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50355</td>\n",
       "      <td>15892491033</td>\n",
       "      <td>52970.7</td>\n",
       "      <td>14301.5</td>\n",
       "      <td>2.04532e+08</td>\n",
       "      <td>21758</td>\n",
       "      <td>7363</td>\n",
       "      <td>94395</td>\n",
       "      <td>40000</td>\n",
       "      <td>129492</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>40000</td>\n",
       "      <td>48681</td>\n",
       "      <td>61758</td>\n",
       "      <td>74177</td>\n",
       "      <td>81444.8</td>\n",
       "      <td>94388</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                dtype     tot       n nmiss miss_perc cardinality  \\\n",
       "emp_no          int32  300024  300024     0         0      300024   \n",
       "no_of_projects  int32  300024  300024     0         0          10   \n",
       "left            int32  300024  300024     0         0           2   \n",
       "Age             int32  300024  300024     0         0          14   \n",
       "Tenure_Years    int32  300024  300024     0         0          28   \n",
       "salary          int32  300024  300024     0         0       50355   \n",
       "\n",
       "                        sum       mean       std          var     iqr  lc_iqr  \\\n",
       "emp_no          76002608740     253322    161828  2.61884e+10  339986 -424973   \n",
       "no_of_projects      1651580    5.50483   2.87207      8.24881       5    -4.5   \n",
       "left                  29867  0.0995487  0.299398    0.0896391       0       0   \n",
       "Age                16476763    54.9181   3.74869      14.0527       6      43   \n",
       "Tenure_Years        6550606    21.8336   5.92815      35.1429       6      11   \n",
       "salary          15892491033    52970.7   14301.5  2.04532e+08   21758    7363   \n",
       "\n",
       "                uc_iqr    min     max       p1       p5      p10      p25  \\\n",
       "emp_no          934973  10001  499999  13001.2  25002.2  40003.3  85006.8   \n",
       "no_of_projects    15.5      1      10        1        1        2        3   \n",
       "left                 0      0       1        0        0        0        0   \n",
       "Age                 67     48      61       49       49       50       52   \n",
       "Tenure_Years        35      1      28        3        8       14       20   \n",
       "salary           94395  40000  129492    40000    40000    40000    40000   \n",
       "\n",
       "                   p50     p75     p90      p95     p99  \n",
       "emp_no          249988  424993  469997   484998  496999  \n",
       "no_of_projects       6       8      10       10      10  \n",
       "left                 0       0       0        1       1  \n",
       "Age                 55      58      60       61      61  \n",
       "Tenure_Years        23      26      28       28      28  \n",
       "salary           48681   61758   74177  81444.8   94388  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def fun_describe(x):\n",
    "\n",
    "    # Records and missing values\n",
    "    n_tot = x.shape[0]\n",
    "    n_count = x.count()\n",
    "    n_miss = x.isna().sum()\n",
    "    n_miss_perc = n_miss / n_tot\n",
    "\n",
    "    # IQR\n",
    "    q1 = x.quantile(0.25)\n",
    "    q3 = x.quantile(0.75)\n",
    "    iqr = q3 - q1\n",
    "    lc_iqr = q1 - 1.5 * iqr\n",
    "    uc_iqr = q3 + 1.5 * iqr\n",
    "\n",
    "    return pd.Series(data=[\n",
    "        x.dtypes, n_tot, n_count, n_miss, n_miss_perc,\n",
    "        x.nunique(),\n",
    "        x.sum(),\n",
    "        x.mean(),\n",
    "        x.std(),\n",
    "        x.var(), iqr, lc_iqr, uc_iqr,\n",
    "        x.min(),\n",
    "        x.max(),\n",
    "        x.quantile(0.01),\n",
    "        x.quantile(0.05),\n",
    "        x.quantile(0.1),\n",
    "        x.quantile(0.25),\n",
    "        x.quantile(0.5),\n",
    "        x.quantile(0.75),\n",
    "        x.quantile(0.90),\n",
    "        x.quantile(0.95),\n",
    "        x.quantile(0.99)\n",
    "    ],\n",
    "        index=[\n",
    "        'dtype', 'tot', 'n', 'nmiss', 'miss_perc',\n",
    "        'cardinality', 'sum', 'mean', 'std', 'var', 'iqr',\n",
    "        'lc_iqr', 'uc_iqr', 'min', 'max', 'p1', 'p5', 'p10',\n",
    "        'p25', 'p50', 'p75', 'p90', 'p95', 'p99'\n",
    "    ])\n",
    "conti_var_df.apply(fun_describe).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, 'emp_no' is having numeric datatype but not a variable to be considered as it has maximum cardinality i.e. primary key, used for identification only. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Categorical Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>dtype</th>\n",
       "      <th>tot</th>\n",
       "      <th>n</th>\n",
       "      <th>nmiss</th>\n",
       "      <th>miss_perc</th>\n",
       "      <th>cardinality</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>dept_no</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>emp_title_id</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>birth_date</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first_name</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_name</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hire_date</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_performance_rating</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_date</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>29867</td>\n",
       "      <td>270157</td>\n",
       "      <td>0.900451</td>\n",
       "      <td>7582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title_id</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dept_name</th>\n",
       "      <td>object</td>\n",
       "      <td>300024</td>\n",
       "      <td>300024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          dtype     tot       n   nmiss miss_perc cardinality\n",
       "dept_no                  object  300024  300024       0         0           9\n",
       "emp_title_id             object  300024  300024       0         0           7\n",
       "birth_date               object  300024  300024       0         0        4750\n",
       "first_name               object  300024  300024       0         0        1276\n",
       "last_name                object  300024  300024       0         0        1638\n",
       "sex                      object  300024  300024       0         0           2\n",
       "hire_date                object  300024  300024       0         0        5434\n",
       "last_performance_rating  object  300024  300024       0         0           5\n",
       "last_date                object  300024   29867  270157  0.900451        7582\n",
       "title_id                 object  300024  300024       0         0           7\n",
       "title                    object  300024  300024       0         0           7\n",
       "dept_name                object  300024  300024       0         0           9"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def fun_obj_describe(x):\n",
    "\n",
    "    # Records and missing values\n",
    "    n_tot = x.shape[0]\n",
    "    n_count = x.count()\n",
    "    n_miss = x.isna().sum()\n",
    "    n_miss_perc = n_miss / n_tot\n",
    "\n",
    "    return pd.Series(data=[\n",
    "        x.dtypes, n_tot, n_count, n_miss, n_miss_perc,\n",
    "        x.nunique()\n",
    "    ],\n",
    "        index=[\n",
    "        'dtype', 'tot', 'n', 'nmiss', 'miss_perc',\n",
    "        'cardinality'\n",
    "    ])\n",
    "\n",
    "cat_var_df.apply(fun_obj_describe).T"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoding columns with high cardinality may effect the model, so all of these categorical cols with high cardinality will be dropped later."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Checking Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVMAAAFYCAYAAADujB+3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAWvklEQVR4nO3df2zUhf3H8Vd7ld8ltUdbrkNWYZOdc2gCG1vEH2DlyrjSmgxrCi6RrWRfN3Xku8XKZkuFRZvlO+cEvjKyH5J+sy1kCcpJSjW6zJrJmDpadxMXKIOFo7RXGn4o0N59vn8YLnS0vWv7Pu4Dez7+Gv18eveivTy9z5XeshzHcQQAGJPsTA8AgGsBMQUAA8QUAAwQUwAwQEwBwAAxBQADOZkecKlo9Izi8cz+S63rr5+kkyc/yuiGi9y0RXLXHrYMzk1bJHftsdpSUJA76Md5ZvpvcnI8mZ6Q4KYtkrv2sGVwbtoiuWtPurcQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMJI1pY2OjFi9erDlz5ujDDz8c9JxYLKaGhgaVlpbq3nvv1Y4dO8yHAoCbJf110nvuuUdf//rXtXLlyiHP2bVrl44cOaKWlhb19vaqsrJSX/nKVzRjxgzTsUN5ufWQdrYeviL3BeDqlJUl5U68TsXTJqtswUzNnT3N9PaTPjOdP3++fD7fsOfs3r1bK1asUHZ2tvLz81VaWqrm5mazkcMhpABS4TjSqY/61HnyY/3fqx+q7WC36e2bvGYaiURUXFyc+LPP59Px48ctbjqpln3/uiL3A+Da8PH5fnk82Wree8T0dl31rlFe75QRf865vlgalgC4VvXHHE2ekKOTZy4M+Q5Qo2ESU5/Pp2PHjmnu3LmSLn+mmqrRvAXfhOs8+uh8/4jvC8B/phxPls6e69f1U8apq+v0iD8/rW/BV1ZWph07digej6unp0evvfaaAoGAxU0nteSLV+aHXACuDRPH5ygWi6tswUzT200a040bN+rOO+/U8ePH9dBDD2nZsmWSpJqaGrW3t0uSKioqNGPGDC1ZskT333+/vv3tb+uGG24wHTqU5QtnqXJhyRW5LwBXr6wsaeqk61R0/UStvPcm85/mZzmOk9m3tr+EG95pv6Agd1RP/dPBTVskd+1hy+DctEVy1x6rLbzTPgCkETEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADCQk8pJHR0dqq2tVW9vr/Ly8tTY2KiSkpIB50SjUT3xxBOKRCLq6+vTl7/8Zf3whz9UTk5KdwEAV7WUnpnW19erurpae/bsUXV1terq6i4754UXXtDs2bO1a9cu7dq1S3/729/U0tJiPhgA3ChpTKPRqMLhsILBoCQpGAwqHA6rp6dnwHlZWVk6e/as4vG4Lly4oL6+PhUVFaVnNQC4TNJr8EgkoqKiInk8HkmSx+NRYWGhIpGI8vPzE+c9/PDDeuSRR7Rw4UJ9/PHHWrlypebNmzeiMV7vlBHOT4+CgtxMT0hw0xbJXXvYMjg3bZHctSedW8xe0GxubtacOXP04osv6uzZs6qpqVFzc7PKyspSvo1o9Iziccdq0qgUFOSqq+t0Rjdc5KYtkrv2sGVwbtoiuWuP1Zahgpz0Mt/n86mzs1OxWEySFIvFdOLECfl8vgHnNTU1afny5crOzlZubq4WL16svXv3jnk4AFwNksbU6/XK7/crFApJkkKhkPx+/4BLfEmaMWOG/vjHP0qSLly4oD/96U/67Gc/m4bJAOA+Kf00f/369WpqalIgEFBTU5MaGhokSTU1NWpvb5ckrVu3Tu+8847Ky8tVWVmpkpIS3X///elbDgAukuU4TmZfpLwEr5kO5KYtkrv2sGVwbtoiuWtPxl8zBQAkR0wBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAAynFtKOjQ1VVVQoEAqqqqtLhw4cHPW/37t0qLy9XMBhUeXm5uru7LbcCgGvlpHJSfX29qqurVVFRoZdeekl1dXXavn37gHPa29u1adMmvfjiiyooKNDp06c1bty4tIwGALdJ+sw0Go0qHA4rGAxKkoLBoMLhsHp6egac9+tf/1qrV69WQUGBJCk3N1fjx49Pw2QAcJ+kMY1EIioqKpLH45EkeTweFRYWKhKJDDjv4MGDOnr0qFauXKn77rtPW7ZskeM46VkNAC6T0mV+KmKxmA4cOKBf/epXunDhgr75zW+quLhYlZWVKd+G1zvFas6YFBTkZnpCgpu2SO7aw5bBuWmL5K496dySNKY+n0+dnZ2KxWLyeDyKxWI6ceKEfD7fgPOKi4tVVlamcePGady4cbrnnnvU1tY2ophGo2cUj2f22WxBQa66uk5ndMNFbtoiuWsPWwbnpi2Su/ZYbRkqyEkv871er/x+v0KhkCQpFArJ7/crPz9/wHnBYFCtra1yHEd9fX16++239bnPfW7MwwHgapDSP41av369mpqaFAgE1NTUpIaGBklSTU2N2tvbJUnLli2T1+vVV7/6VVVWVuozn/mMvva1r6VvOQC4SJbjop8ScZk/kJu2SO7aw5bBuWmL5K49Gb/MBwAkR0wBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAykFNOOjg5VVVUpEAioqqpKhw8fHvLcQ4cO6dZbb1VjY6PVRgBwvZRiWl9fr+rqau3Zs0fV1dWqq6sb9LxYLKb6+nqVlpaajgQAt0sa02g0qnA4rGAwKEkKBoMKh8Pq6em57Nyf//znuvvuu1VSUmI+FADcLCfZCZFIREVFRfJ4PJIkj8ejwsJCRSIR5efnJ8774IMP1Nraqu3bt2vLli2jGuP1ThnV51krKMjN9IQEN22R3LWHLYNz0xbJXXvSuSVpTFPR19enJ598Uk8//XQiuqMRjZ5RPO5YTBq1goJcdXWdzuiGi9y0RXLXHrYMzk1bJHftsdoyVJCTxtTn86mzs1OxWEwej0exWEwnTpyQz+dLnNPV1aUjR45ozZo1kqRTp07JcRydOXNGGzZsGPN4AHC7pDH1er3y+/0KhUKqqKhQKBSS3+8fcIlfXFysvXv3Jv78/PPP66OPPtLjjz+entUA4DIp/TR//fr1ampqUiAQUFNTkxoaGiRJNTU1am9vT+tAALgapPSa6ezZs7Vjx47LPr5t27ZBz3/kkUfGtgoArjL8BhQAGCCmAGCAmAKAAWIKAAaIKQAYIKYAYICYAoABYgoABogpABggpgBggJgCgAFiCgAGiCkAGCCmAGCAmAKAAWIKAAaIKQAYIKYAYICYAoABYgoABogpABggpgBggJgCgAFiCgAGiCkAGCCmAGCAmAKAAWIKAAaIKQAYIKYAYICYAoABYgoABogpABggpgBggJgCgAFiCgAGiCkAGCCmAGCAmAKAAWIKAAaIKQAYIKYAYICYAoABYgoABogpABggpgBggJgCgIGcVE7q6OhQbW2tent7lZeXp8bGRpWUlAw4Z/Pmzdq9e7c8Ho9ycnK0du1a3XHHHenYDACuk1JM6+vrVV1drYqKCr300kuqq6vT9u3bB5wzd+5crV69WhMnTtQHH3ygVatWqbW1VRMmTEjLcABwk6SX+dFoVOFwWMFgUJIUDAYVDofV09Mz4Lw77rhDEydOlCTNmTNHjuOot7c3DZMBwH2SxjQSiaioqEgej0eS5PF4VFhYqEgkMuTn7Ny5UzNnztT06dPtlgKAi6V0mT8Sf/7zn/Xcc8/pl7/85Yg/1+udYj1nVAoKcjM9IcFNWyR37WHL4Ny0RXLXnnRuSRpTn8+nzs5OxWIxeTwexWIxnThxQj6f77Jz33vvPX3/+9/Xli1bNGvWrBGPiUbPKB53Rvx5lgoKctXVdTqjGy5y0xbJXXvYMjg3bZHctcdqy1BBTnqZ7/V65ff7FQqFJEmhUEh+v1/5+fkDzmtra9PatWv1s5/9TJ///OfHPBgAriYp/TvT9evXq6mpSYFAQE1NTWpoaJAk1dTUqL29XZLU0NCgc+fOqa6uThUVFaqoqNCBAwfStxwAXCSl10xnz56tHTt2XPbxbdu2Jf7373//e7tVAHCV4TegAMAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMJCTykkdHR2qra1Vb2+v8vLy1NjYqJKSkgHnxGIxbdy4UW+++aaysrK0Zs0arVixIh2bL/Nf//OGzvc5V+S+AFz9KheWaPnCWaa3mdIz0/r6elVXV2vPnj2qrq5WXV3dZefs2rVLR44cUUtLi373u9/p+eef17/+9S/TsYMhpABGamfrYb3cesj0NpPGNBqNKhwOKxgMSpKCwaDC4bB6enoGnLd7926tWLFC2dnZys/PV2lpqZqbm03HDoaQAhiNln22T/aSXuZHIhEVFRXJ4/FIkjwejwoLCxWJRJSfnz/gvOLi4sSffT6fjh8/PqIxXu+UEZ0PAKN1ri+mgoJcs9tL6TXTKyUaPaN4nGeaANJvwnUedXWdHvHnDRXgpJf5Pp9PnZ2disVikj75QdOJEyfk8/kuO+/YsWOJP0ciEU2fPn3EQ0dq/HVZab8PANeeJV+cYXp7SWPq9Xrl9/sVCoUkSaFQSH6/f8AlviSVlZVpx44disfj6unp0WuvvaZAIGA6djD/+9+LCCqAEUnHT/OzHMdJel198OBB1dbW6tSpU5o6daoaGxs1a9Ys1dTU6NFHH9UXvvAFxWIxPfXUU3rrrbckSTU1NaqqqhrRGDdc5hcU5I7qqX86uGmL5K49bBmcm7ZI7tpjtWWoy/yUYnqlENOB3LRFctcetgzOTVskd+1Jd0z5DSgAMEBMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADrvrd/Oxsd/wmk1t2SO7aIrlrD1sG56Ytkrv2pHOLq/7RPgBcrbjMBwADxBQADBBTADBATAHAADEFAAPEFAAMEFMAMEBMAcAAMQUAA9d0TBsbG7V48WLNmTNHH374YeLjf/jDH3TfffepvLxcq1at0tGjRxPH3njjDVVWVqqiokLl5eVqaWlJHDt//rzq6+u1ZMkSlZeX68knn8zonuGOpWPLcMc6OjpUVVWlQCCgqqoqHT58OCNbTp48qZqaGgUCAZWXl+s73/mOenp6MvZ1uWjTpk2X3WYm9ozlMWy9ZSyP3+G+z3/961+1fPlyBQIBrV69WtFoNPF5oz2WEucatm/fPufYsWPOokWLnAMHDjiO4zi9vb3Ol770JefQoUOO4zjOzp07ndWrVzuO4zjxeNyZP39+4ty///3vzm233ebEYjHHcRxnw4YNzo9+9CMnHo87juM4XV1dGduTbKv1luGOOY7jPPjgg87OnTsTxx588MG0fV2GO3by5Enn7bffTtz2M8884zzxxBMZ2XLR+++/73zjG99w7r777sRtZmrPWB7DllvG+vgd6vscj8ed0tJSZ9++fY7jOM7mzZud2traxH2O5liqrulnpvPnz5fP5xvwsX/+85+aNm2abrzxRknSXXfdpdbW1sR/1bKzs3X69Cf/p1unT59WYWGhsrOzdfbsWe3cuVOPPfaYsrI+ebOEadOmZWxPsmPWW4Y7Fo1GFQ6HFQwGJUnBYFDhcDjlZ4SWW/Ly8rRgwYLE7dx22206duxYSjust0jShQsX9NRTT6m+vj7xuBkJyz1jfQxbf23G8vgd6vvc3t6u8ePHa/78+ZKkBx54QM3NzZI06mOpctW7Rl0JN954o7q7u9XW1qa5c+dq165dkqRIJKL8/Hz99Kc/1cMPP6xJkybp7Nmz2rp1qyTp6NGjysvL06ZNm7R3715NnjxZjz32WOKLf6X3ZGVlDXksHVuGO+Y4joqKiuTxeCRJHo9HhYWFib/Dldxy6f3F43H95je/0eLFi0f9NRnrlueee07Lly/XDTfcMKYNFns8Ho/5Y3gsXxurx++l3+dIJKLi4uLEsfz8fMXjcfX29o76WF5eXko7/uNimpubq2effVZPP/20zp8/rzvvvFNTp05VTk6O+vv7tXXrVm3ZskXz5s3TO++8o7Vr1+qVV15Rf3+/jh49qptvvlmPP/649u/fr29961t69dVXNWXKlCu+Z/z48UMemzx5svmW4Y719fWN+u9vveVSGzZs0KRJk7Rq1aqMbHnvvffU3t6u733ve2O6f6s9fX195o/h0W4Z7rE90sfvpd/nV199dVR/DxMjelHgKnXpazz/rqury7nlllucs2fPOm1tbc7SpUsHHC8rK3P279/vRKNR5+abb0681uQ4jrN06VKnra0tI3uGO5aOLcMd6+7udubNm+f09/c7juM4/f39zrx585xoNHrFt1z0zDPPOA899JBz/vz5EW2w3LJ161bn9ttvdxYtWuQsWrTI8fv9zsKFC50333wzI3usHsMWW6wev//+fd6/f7+zbNmyxPFoNOrceuutYzqWqmv6NdOhdHV1Sfrk8uAnP/mJHnjgAU2aNEnTp0/X8ePHdejQIUnSwYMH1d3drZkzZyo/P18LFizQW2+9JemTn15Ho1F9+tOfzsie4Y6lY8twx7xer/x+v0KhkCQpFArJ7/eP+hJ/LFsk6dlnn9X777+vzZs3a9y4cWPaMJYta9asUWtrq15//XW9/vrrmj59un7xi19o4cKFGdmTrsfwaLZYPH4H+z7fcsstOnfunP7yl79Ikn77299q6dKlYzqWqmv6zaE3btyolpYWdXd36/rrr1deXp5eeeUV/eAHP9C7776rvr4+3X777Vq3bp3Gjx8vSXr55Ze1bdu2xAv0jz76qEpLSyV98rrpunXr1Nvbq5ycHH33u9/VXXfdlbE9wx1Lx5bhjh08eFC1tbU6deqUpk6dqsbGRs2aNeuKb/nHP/6hYDCokpISTZgwQZI0Y8YMbd68OSNfl0stXrxYL7zwgm666aaUtqRjz1gew9ZbxvL4He77/O6776q+vl7nz5/Xpz71Kf34xz9O/KBttMdScU3HFACulP/Iy3wAsEZMAcAAMQUAA8QUAAwQUwAwQEwBwAAxBQADxBQADPw/Ul0mJfsp/lQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFcCAYAAAAZGw+kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAV40lEQVR4nO3ce2zVd/3H8Vd7ujomFThnbXdKmUg3uhM3FN0fM7LpoFu5HFKmg8Y6Y0Y4LnPggpdZYOsFvKz6hzroNOIsYKdiVxjhpBaCC5mo4JQ5Ct1wzGJZOJRyDhXYZHTnfH9/kDXUAj2U8+37t9PnIyGhPZ9z3p/vgT5z+j2XDMdxHAEATGRabwAARjIiDACGiDAAGCLCAGCICAOAISIMAIYGjXBdXZ2mT5+u4uJi/fOf/7zomng8rtraWpWUlOiee+5RU1NTyjcKAOlo0AjPmDFDzz77rMaPH3/JNVu3blVnZ6e2b9+ujRs3avXq1XrzzTdTulEASEeDRvj222+X3++/7JqWlhbNnz9fmZmZ8nq9KikpUWtra8o2CQDpKiXnhCORiAoKCvq+9vv9OnbsWCpuGgDSGk/MAYChrFTciN/v19GjRzVlyhRJAx8ZJysaPaNEYng/yiI3N0fd3aeHdeZImzuSjpW5I2NuKqXkkfDMmTPV1NSkRCKhWCymHTt2qLS0NBU3DQBpbdAIf+c739Fdd92lY8eO6cEHH9ScOXMkSaFQSG1tbZKksrIyFRYW6t5779WCBQv0yCOPaMKECe7uHADSQMb/p4+y5HREes4dScfK3JExN5V4Yg4ADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMZSWzqKOjQ5WVlerp6dHYsWNVV1eniRMn9lsTjUa1bNkyRSIR9fb26o477tDjjz+urKykRgDAiJTUI+Hq6mpVVFRo27ZtqqioUFVV1YA1P/vZz1RUVKStW7dq69atOnDggLZv357yDQNAOhk0wtFoVO3t7QoGg5KkYDCo9vZ2xWKxfusyMjL01ltvKZFI6Ny5c+rt7VV+fr47uwaANDFohCORiPLz8+XxeCRJHo9HeXl5ikQi/dZ99atfVUdHh6ZNm9b355Of/KQ7uwaANJGyE7atra0qLi7W+vXr9dZbbykUCqm1tVUzZ85M+jZ8vtGp2s4Vyc3NYW4azmQuc98PBo2w3+9XV1eX4vG4PB6P4vG4jh8/Lr/f329dY2Ojvve97ykzM1M5OTmaPn269uzZc0URjkbPKJFwrvworkJubo66u08P68yRNnckHStzR8bcVBr0dITP51MgEFA4HJYkhcNhBQIBeb3efusKCwv14osvSpLOnTunv/zlL7r55ptTulkASDdJvTqipqZGjY2NKi0tVWNjo2prayVJoVBIbW1tkqTly5fr73//u+bOnat58+Zp4sSJWrBggXs7B4A0kNQ54aKiIjU1NQ34/tq1a/v+fuONN6qhoSF1OwOAEYB3zAGAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4ChpCLc0dGh8vJylZaWqry8XIcPH77oupaWFs2dO1fBYFBz587ViRMnUrlXAEg7Wcksqq6uVkVFhcrKyrRlyxZVVVVpw4YN/da0tbVpzZo1Wr9+vXJzc3X69GllZ2e7smkASBeDPhKORqNqb29XMBiUJAWDQbW3tysWi/Vbt27dOi1cuFC5ubmSpJycHH3gAx9wYcsAkD4GjXAkElF+fr48Ho8kyePxKC8vT5FIpN+6N954Q0eOHNEXv/hF3XfffXr66aflOI47uwaANJHU6YhkxONxHTx4UA0NDTp37pwWLVqkgoICzZs3L+nb8PlGp2o7VyQ3N4e5aTiTucx9Pxg0wn6/X11dXYrH4/J4PIrH4zp+/Lj8fn+/dQUFBZo5c6ays7OVnZ2tGTNmaN++fVcU4Wj0jBKJ4X30nJubo+7u08M6c6TNHUnHytyRMTeVBj0d4fP5FAgEFA6HJUnhcFiBQEBer7ffumAwqF27dslxHPX29mr37t265ZZbUrpZAEg3Sb1EraamRo2NjSotLVVjY6Nqa2slSaFQSG1tbZKkOXPmyOfzafbs2Zo3b55uuukm3X///e7tHADSQIbz/+jZM05HpOfckXSszB0Zc1OJd8wBgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaSinBHR4fKy8tVWlqq8vJyHT58+JJr//Wvf+ljH/uY6urqUrVHAEhbSUW4urpaFRUV2rZtmyoqKlRVVXXRdfF4XNXV1SopKUnpJgEgXQ0a4Wg0qvb2dgWDQUlSMBhUe3u7YrHYgLU///nP9dnPflYTJ05M+UYBIB1lDbYgEokoPz9fHo9HkuTxeJSXl6dIJCKv19u37rXXXtOuXbu0YcMGPf3000PajM83ekjXu1q5uTnMTcOZzGXu+8GgEU5Gb2+vnnjiCX3/+9/vi/VQRKNnlEg4qdhS0nJzc9TdfXpYZ460uSPpWJk7Muam0qAR9vv96urqUjwel8fjUTwe1/Hjx+X3+/vWdHd3q7OzU1/5ylckSadOnZLjODpz5oxWrVqV0g0DQDoZNMI+n0+BQEDhcFhlZWUKh8MKBAL9TkUUFBRoz549fV+vXr1ab7/9tr797W+7s2sASBNJvTqipqZGjY2NKi0tVWNjo2prayVJoVBIbW1trm4QANJZUueEi4qK1NTUNOD7a9euvej6JUuWXN2uAGCE4B1zAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYIgIA4AhIgwAhogwABgiwgBgiAgDgCEiDACGiDAAGCLCAGCICAOAISIMAIaIMAAYIsIAYCgrmUUdHR2qrKxUT0+Pxo4dq7q6Ok2cOLHfmvr6erW0tMjj8SgrK0tLly7VnXfe6caeASBtJBXh6upqVVRUqKysTFu2bFFVVZU2bNjQb82UKVO0cOFCjRo1Sq+99poeeOAB7dq1S9dee60rGweAdDDo6YhoNKr29nYFg0FJUjAYVHt7u2KxWL91d955p0aNGiVJKi4uluM46unpcWHLAJA+Bn0kHIlElJ+fL4/HI0nyeDzKy8tTJBKR1+u96HWef/553XjjjbrhhhuuaDM+3+grWp8qubk5zE3Dmcxl7vtBUqcjrsRf//pX/eQnP9Evf/nLK75uNHpGiYST6i1dVm5ujrq7Tw/rzJE2dyQdK3NHxtxUGvR0hN/vV1dXl+LxuCQpHo/r+PHj8vv9A9a+/PLL+ta3vqX6+npNmjQppRsFgHQ0aIR9Pp8CgYDC4bAkKRwOKxAIDDgVsW/fPi1dulRPPfWUPvrRj7qzWwBIM0m9TrimpkaNjY0qLS1VY2OjamtrJUmhUEhtbW2SpNraWp09e1ZVVVUqKytTWVmZDh486N7OASANJHVOuKioSE1NTQO+v3bt2r6/Nzc3p25XADBC8I45ADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ0QYAAwRYQAwRIQBwBARBgBDRBgADBFhADBEhAHAEBEGAENEGAAMEWEAMJRUhDs6OlReXq7S0lKVl5fr8OHDA9bE43HV1taqpKRE99xzj5qamlK9VwBIO0lFuLq6WhUVFdq2bZsqKipUVVU1YM3WrVvV2dmp7du3a+PGjVq9erXefPPNlG8YANJJ1mALotGo2tvb1dDQIEkKBoNatWqVYrGYvF5v37qWlhbNnz9fmZmZ8nq9KikpUWtrqxYtWpT0ZjIzM4ZwCFePuek5k7nMfT8YNMKRSET5+fnyeDySJI/Ho7y8PEUikX4RjkQiKigo6Pva7/fr2LFjV7SZceM+eEXrU8XnG83cNJzJXOa+H/DEHAAYGjTCfr9fXV1disfjks4/AXf8+HH5/f4B644ePdr3dSQS0Q033JDi7QJAehk0wj6fT4FAQOFwWJIUDocVCAT6nYqQpJkzZ6qpqUmJREKxWEw7duxQaWmpO7sGgDSR4TiOM9iiN954Q5WVlTp16pQ+9KEPqa6uTpMmTVIoFNLXvvY13XbbbYrH41q5cqX+9Kc/SZJCoZDKy8tdPwAAeD9LKsIAAHfwxBwAGCLCAGCICAOAISIMAIYGfcfcUJw8eVKPPfaYOjs7lZ2drQ9/+MNauXKlvF6v/vGPf6iqqkrvvPOOxo8frx/+8Ify+XySNOTL3Jz7jW98Q3v27FF3d7f27t2rD35w4Lv6Uj23o6NDVVVV6u7uVlZWlm677TZVV1fr2muvdXVuIpHQF77wBf33v/+VJOXm5qq2tlaFhYWu3sfvWbZsmTZt2nTR+9mNucXFxZo8ebIyM88/FvnBD36g4uJi1+f29PRo5cqVOnDggLKysjRr1iwtXrzY1bl79+5VbW1t3+1Ho1Hl5uZq8+bNrh/vc889p/Xr1yszM1Mej0fLly/X7bff7vrc5uZmrVu3TolEQhMmTNCTTz6psWPHXvXMyzUhmf/nAzguOHnypLN79+6+r5988kln2bJlTiKRcEpKSpyXXnrJcRzHqa+vdyorKx3HcYZ8mZtzHcdx/vznPzsnTpxwJk+e7Jw5c2ZYjvfIkSPOgQMHHMdxnHg87jz66KPOmjVrhuV4T5061ff3devWOY888ojrMx3Hcf7whz84y5Ytu+T97Mbcy/2bujn3oYcechoaGvq+7u7uHpa5F3r44YedX/ziF67PjcViztSpU/uOcceOHc6sWbNcn3vo0CFn2rRpTjQa7bvsiSeeuOqZjnPpJlzJ/X8hVyL8v1pbW50vf/nLziuvvOLMmTOn7/vRaNT5+Mc/7jiOM+TL3Jx7oWR+YN2Y6ziO88wzzzjLly8f9rlr1qxxlixZ4vrMWCzm3Hfffc6pU6eSvp9TMfdK/k1TNbejo8O5++67nUQiMaxzL3TixAlnypQpF41/qudGo1Fn6tSpzuHDhx3HcZzNmzc7Dz74oOtzW1panFAo1HfZ/v37B21GMjMv9L//f4baKVdOR1wokUjoN7/5jaZPnz7gQ368Xq8SiYR6enqGfNmFv16keu6lbtvt471w7tmzZ9Xc3Kyvf/3rwzY3FAqpvb1d48aN0zPPPOP6zJUrV2rJkiXKycm57DG6caxf+tKXFI/Hddddd2nJkiXKzs52de6hQ4eUn5+vFStW6NVXX9X111+vxx57TDfffPOwHK8kPf/88/r0pz+t66+/3vX72ev1qqamRvPmzdOYMWOUSCT0q1/9yvW5t9xyi/bv368jR46osLBQ4XBYb7/99iV/rpOdebkmDPV6rj8xt2rVKl133XV64IEH3B6VdnPfffddLV26VHfccYdmzJgxbHPXrl2rP/7xj5ozZ45++tOfujrz97//va655hrdfffdSV8nVce6c+dObdq0Sc8++6wOHTqk+vp61+fG43G98sor+tznPqfNmzdr/vz5evjhh12fe6FNmzbp85///KDrUjH3zJkz+vWvf63m5mbt3LlTlZWVWrx4sZzLvEcsFXM/8pGPaMWKFVq6dKkWLFjQF8GsrIs/7rTqheRyhOvq6vTvf/9bP/7xj5WZmTngQ35isZgyMjI0duzYIV/m5lyr45XO/7B+85vf1JgxY/T4448P29z3ZGZm6v7779eWLVtcnblnzx7t3r1b06dP1/Tp0yWd/8zqQ4cOuX6s730I1ejRozV//nzt3bvX9fu4oKBAfr+/74mpe++9V93d3YrFYq4fr3T+iaOenh595jOfueSxpnLurl27lJOTo0mTJkmSZs+erc7OTp08edL1450zZ46ee+45NTU16VOf+pTy8/M1evTAj768kpmXM9TruRbhH/3oR9q/f7/q6+v7fsW79dZbdfbsWf3tb3+TJP32t7/VrFmzruoyN+daHW8ikVBlZaU8Ho+++93vKiPj0h9cncq5sVis3w9Ha2vrgFcLpHpmTU2NXnzxRb3wwgt64YUXJJ3/kKibbrrJ1bn/+c9/dPbsWUnnf+PYtm2bAoGA6/fxrbfequuuu06vv/66JOmll17SmDFjNG7cOFfnvqe5uVllZWWXfESY6rmFhYV69dVXFY1GJUm7d+/W6NGjh+V4u7u7JUnvvPOOnnrqKS1cuPCqZ17OUK/nymdHvP766woGg5o4cWLfy6oKCwtVX1+vvXv3qrq6ut9LON47NzXUy9ycu3jxYu3bt09dXV3Ky8vT5MmTB5wnTfXcnTt36qGHHur38qlPfOITqq6udnXuwYMHtWzZMvX29kqSxo8frxUrVmjChAmu3scXKi4uvuhL1FI99+WXX1ZVVZUyMjL07rvvaurUqVq+fLnrcyWpra1NtbW1OnfunEaNGqUVK1ZoypQprs89e/aspk2bpo0bN6qoqGjAfe/W3IaGBv3ud7/TNddco+zsbFVWVg54iZobcxctWqSjR4+qt7dXs2fP1qOPPtr383Q1My/XhGT/n1+ID/ABAEO8Yw4ADBFhADBEhAHAEBEGAENEGAAMEWEAMESEAcAQEQYAQ/8HmkMV0lZ7UFYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVwAAAFcCAYAAACEFgYsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAfrUlEQVR4nO3de3BV1f338c8+OzcIwZAYwomI/PASUxVFbKl94FerQFATSKdinCCjDxpm6igdxrYiCuFiW+LYsV7w1rFWi9oOnQ5ISgEVbMUHEC0KNiA1gIgcSEhA7iTZez1/xBxyIwTIWUlO3q8/Ss45a+/9/caez9p7JTnbMcYYAQAiLtDRBQBAd0HgAoAlBC4AWELgAoAlBC4AWELgAoAlMR1dQHurrDws3z+733Tr06en9u8/2s4VdT7doc/u0KNEn51RWlrSKV/jDLeBmBi3o0uwojv02R16lOizqyFwAcASAhcALCFwAcASAhcALCFwAcASAhcALCFwAcASAhcALLHyl2bFxcVavny5vv76ay1ZskSXXXZZszGe5+mxxx7T+++/L8dxNHnyZI0fP95GeZo0b6WV4wDRKNZ1dPEF52nMsAEafPH5emv1Nq1Yv0vHq2uVEBejqy9O0f7D1dr3zXElxAb0zdEaHTpa02gf8bEBxcW6uuD8RPXpFadPy6rC24/+bn9dfXk//WXFFu375rjOPy8hfKyGNpbt07J1O1sd09EcG3d8+Oijj3TBBRdowoQJeuGFF1oM3EWLFmnJkiX6/e9/rwMHDigvL09vvPGG+vfvf0bHOtM/7SVsgXMX5wZ0XlKcLsnorXVbyuXIUcCRPN/IN1LPeFc9E2K075sTp9xHICDFugGdqPEVcCQ34Mg3km+MesS56t0rXnExAVXX+vI8XxNGXRYO1I1l+/T621vluoFTjrGlw/+097rrrlMwGGx1zNKlSzV+/HgFAgGlpKRo5MiRWrZsmY3yAJyjas+X6wa0bnOFHDlyA44cx1H9qc+xak+HjtW2ug/fl07U+JIkI8lx6vZjTN328bGuHMdRfKwr1w1o2bqd4W2Xrdsp1w20OqYz6DQfXhMKhZSRkRF+HAwGtWfPnjPeT2pqr/YsC0AbJSbEyDdGcW5d2EpS/fWzMZLn+W3elzEK76P+cWzMyfPDGNfR/sPV4bPJqsPVSuoR02ibpmM6g04TuO3lXD4tDMDZO3K8VgHHkedLbqDuPeg49eEpuW5AtZ7Xpn3VbWcaPa6pPRnYJ2o89ekVp4qKQ5KklF5xOnCkWvGx7inH2NLhSwptEQwGtXv37vDjUCikfv36dWBFANoqzg3I83wNy0qTkZHnGxljVH++2SPOVVKP1s/vAoG6H55JkqO6wPV8I8ep2/5EjSdjjE7UePI8X2OGDQhvO2bYAHme3+qYzqDTBO6YMWO0cOFC+b6vqqoqvfPOO8rOzo74cf8w7caIHwOIZrGuo0EX9NaEUZepcOyVGvuDgYqPdeX5RglxMbr+O301ID1JkqP+5/dUUs/YZvuIjw0oMSFW/xPsreu/01cJcTHyfKP4WFfj/s9A/WLid5WcGKejx2uVnBjX7Idhgy8+XxNGXdbqmM7Aym8pPPbYY1qxYoX27dunPn36KDk5WX//+99VWFioKVOm6KqrrpLneZozZ44++OADSVJhYaHy8/PP+FjnsqSQlpZk/fKjI3SHPrtDjxJ9dkatLSlYCVybCNzT6w59doceJfrsjLrEGi4ARDsCFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwJIYWwfavn27pk2bpgMHDig5OVnFxcUaOHBgozGVlZV6+OGHFQqFVFNTo+9///t69NFHFRNjrUwAiBhrZ7hFRUUqKCjQ8uXLVVBQoJkzZzYb88ILL+jiiy/WkiVLtGTJEv3nP//RihUrbJUIABFlJXArKytVWlqqnJwcSVJOTo5KS0tVVVXVaJzjODpy5Ih831d1dbVqamqUnp5uo0QAiDgr1+qhUEjp6elyXVeS5Lqu+vbtq1AopJSUlPC4++67Tw888ICGDx+uY8eOacKECRo6dOgZHSs1tdc51ZqWlnRO23cV3aHP7tCjRJ9dSadaHF22bJkyMzP16quv6siRIyosLNSyZcs0ZsyYNu+jsvKwfN+c1fHT0pJUUXHorLbtSrpDn92hR4k+O6PWJgYrSwrBYFB79+6V53mSJM/zVF5ermAw2GjcggULNHbsWAUCASUlJenGG2/UunXrbJQIABFnJXBTU1OVlZWlkpISSVJJSYmysrIaLSdIUv/+/fWvf/1LklRdXa01a9bo0ksvtVEiAESctd9SmDVrlhYsWKDs7GwtWLBAs2fPliQVFhZq06ZNkqTp06fr448/Vm5urvLy8jRw4EDdfvvttkoEgIhyjDFnt+DZSbGGe3rdoc/u0KNEn51Rh6/hAgAIXACwhsAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEsIXACwhMAFAEusBe727duVn5+v7Oxs5efna8eOHS2OW7p0qXJzc5WTk6Pc3Fzt27fPVokAEFExtg5UVFSkgoICjRs3TosXL9bMmTP12muvNRqzadMmPfvss3r11VeVlpamQ4cOKS4uzlaJABBRVs5wKysrVVpaqpycHElSTk6OSktLVVVV1WjcH//4R02aNElpaWmSpKSkJMXHx9soEQAizsoZbigUUnp6ulzXlSS5rqu+ffsqFAopJSUlPK6srEz9+/fXhAkTdPToUY0aNUo//elP5ThOm4+VmtrrnGpNS0s6p+27iu7QZ3foUaLPrsTakkJbeJ6nzz//XK+88oqqq6t17733KiMjQ3l5eW3eR2XlYfm+Oavjp6UlqaLi0Flt25V0hz67Q48SfXZGrU0MVpYUgsGg9u7dK8/zJNUFa3l5uYLBYKNxGRkZGjNmjOLi4tSrVy/ddNNN2rhxo40SASDirARuamqqsrKyVFJSIkkqKSlRVlZWo+UEqW5td/Xq1TLGqKamRmvXrtXll19uo0QAiDhrvxY2a9YsLViwQNnZ2VqwYIFmz54tSSosLNSmTZskSbfeeqtSU1N1yy23KC8vT5dccoluu+02WyUCQEQ5xpizW/DspFjDPb3u0Gd36FGiz86ow9dwAQAELgBYQ+ACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCVtDtyXX365xedfeeWVdisGAKJZmwN3/vz5LT7//PPPt1sxABDNTnvX3jVr1kiSfN/X2rVr1fAGEbt27VJiYmLkqgOAKHLawH3kkUckSSdOnND06dPDzzuOo7S0ND366KORqw4AokirgbtlyxatXLlSkvTLX/5Sjz/+uJWiACAatbqGW1BQEP76k08+iXgxABDNWj3D7d27t1atWqVLLrlEFRUV+uqrr1ocd+GFF0akOACIJq0G7iOPPKJf//rX2r17t3zf16hRo5qNcRxHmzdvjliBABAtWg3cUaNGhUN2yJAh2rBhg5WiACAatfn3cNetWyep7tfDysvLI1YQAESrNgfu8ePH9eCDD2rw4MEaPXq0JOndd9/Vk08+GbHiACCatDlwi4qK1KtXL61cuVKxsbGS6pYZ/vGPf0SsOACIJqf9w4d6a9as0fvvv6/Y2Fg5jiNJSklJUWVlZcSKA4Bo0uYz3KSkJO3fv7/Rc7t371ZaWlq7FwUA0ajNgTt+/HhNmTJFa9eule/72rBhgx566CHdcccdkawPAKJGm5cUCgsLFRcXpzlz5qi2tlbTp0/XHXfcobvuuiuS9QFA1Gg1cOs/KaxeZmamZsyY0WzM9ddf3/6VAUCUOe1fmp2K4zgyxshxHL377rvtXhgARJtWA7f+k8IAAOeOe5oBgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCXWAnf79u3Kz89Xdna28vPztWPHjlOO3bZtm66++moVFxfbKg8AIs5a4BYVFamgoEDLly9XQUGBZs6c2eI4z/NUVFSkkSNH2ioNAKywEriVlZUqLS1VTk6OJCknJ0elpaWqqqpqNvall17SDTfcoIEDB9ooDQCssRK4oVBI6enpcl1XkuS6rvr27atQKNRo3JYtW7R69WrdfffdNsoCAKvafIudSKupqdGMGTP0m9/8JhzMZyM1tdc51ZGWlnRO23cV3aHP7tCjRJ9diZXADQaD2rt3rzzPk+u68jxP5eXlCgaD4TEVFRXauXOnJk+eLEk6ePCgjDE6fPiw5s6d2+ZjVVYelu+bs6ozLS1JFRWHzmrbrqQ79NkdepToszNqbWKwEripqanKyspSSUmJxo0bp5KSEmVlZSklJSU8JiMjQ+vWrQs/fuaZZ3T06FE99NBDNkoEgIiz9lsKs2bN0oIFC5Sdna0FCxZo9uzZkuruBrxp0yZbZQBAh3GMMWd3/d1JsaRwet2hz+7Qo0SfnVFrSwr8pRkAWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWELgAoAlBC4AWBJj60Dbt2/XtGnTdODAASUnJ6u4uFgDBw5sNGb+/PlaunSpXNdVTEyMpk6dqhEjRtgqEQAiylrgFhUVqaCgQOPGjdPixYs1c+ZMvfbaa43GDB48WJMmTVKPHj20ZcsW3XnnnVq9erUSEhJslQkAEWNlSaGyslKlpaXKycmRJOXk5Ki0tFRVVVWNxo0YMUI9evSQJGVmZsoYowMHDtgoEQAizkrghkIhpaeny3VdSZLruurbt69CodApt1m0aJEGDBigfv362SgRACLO2pLCmfjwww/11FNP6Q9/+MMZb5ua2uucjp2WlnRO23cV3aHP7tCjRJ9diZXADQaD2rt3rzzPk+u68jxP5eXlCgaDzcZu2LBBv/jFL/Tcc89p0KBBZ3ysysrD8n1zVnWmpSWpouLQWW3blXSHPrtDjxJ9dkatTQxWlhRSU1OVlZWlkpISSVJJSYmysrKUkpLSaNzGjRs1depUPf3007riiitslAYA1jjGmLM7HTxDZWVlmjZtmg4ePKjevXuruLhYgwYNUmFhoaZMmaKrrrpKP/nJT/T1118rPT09vN3jjz+uzMzMNh+HM9zT6w59doceJfrsjFo7w7UWuLYQuKfXHfrsDj1K9NkZdfiSAgCAwAUAawhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUASwhcALCEwAUAS2I6uoDO4oNNIX22o1TV1bUKOI4cRwoEnJNfO46cgKOAIzlOk+cdR4FA8+cDDcbX789R3fPhMY6+3W/D5xw5gQb7bsNxGtbS9PlG+wg4Mq6r/QePf7u/JrU03HeT/QI4NwSuJGOM/vpemb45Ut3RpXRajnTawHeahHazseEJp8GYhhNCwGm8j4AjR00mvkCTSaXJ8/XbJibG6cTxmkbHb+k4DSe1+omv0STY5Pit1dX0OM0ntOa1hidynaL/cI0tT67oWrhN+rc276jSll3f6PDRahnfyBjJN0a+Ofl1+F/fyDd1QR1+3m88tm5c/Zim/zbex8n9Nxjjt/Bc/T717XO+UVT9x8MZaxj+4X9bm0hamijV+GqspSuspvtqeoxmk0ULE2+zuppOXi1MTvWTfK/EBB07Vt3CBH5ysmvUQ0DNriabTvonrxobfh+k5F5xSkyIO+v/Jq3dJp3AbSAtLUkVFYfauaLIMqYudH3ftBjULU0QffokqqrqcHhsS5NGs/D3ffnfHscP78+vO4bfeNLw/KYTSwt1Ndmm7nW/7jnVTSz1tfkN9heeaIyR1+Brv8mkFeMGdKK6tm5bNe2v4cRY32PDvht/D+r7MFKLk2P9vo1OPkbXFesGNPP/flcXnJ94Vtu3FrgsKXRxTv1lr9v2y8u01ES5vh+5ojqB002eZ3Y17oQntvAlhVP3RcNwNd8OMI0micYB7XnNJ8P6kPeaTES+7ze6OmppskvsGa+Dh45/O75+ovIb7KPhZGWaTLJNJ2mF99P86ksnxzeYiJteoTW9MjSmhcmrlSu75serexwIOKqp9Zvtz/hGvuqv9ppv33DCbatAwJEbiMxyDYGLbunMzkLrBjvh/zn5RaPgbjjAPYfizsCZXpW1daKp//60bZKpG2RMXfcNv7VGLXyvG+zUOM1fr7/obnjs5OSe2r//iBo8dXI7c3KbxrWfbMJI8hpedTW4YgtfoXz7b3KvOKWe16Plb8w5InCBbuRMlzvaNsk0+iIizusVr+pjZ/5D7TP9uWKkl4MIXABRq7Otp/OHDwBgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgCYELAJYQuABgibXA3b59u/Lz85Wdna38/Hzt2LGj2RjP8zR79myNHDlSo0aN0sKFC22VBwARZ+0WO0VFRSooKNC4ceO0ePFizZw5U6+99lqjMUuWLNHOnTu1YsUKHThwQHl5ebr++uvVv3//iNY2ad7KiO4fQGQEJDmBk3dVNkYKOI7iYx1d1K+3Lh+QrC07D2jfN8eVEBvQ8Rpf+w+dkO8bOY6UkdpTt/3oEg2++HxtLNunZet2at83x3X+eQkaM2yABl98frvXG3GVlZUqLS1VTk6OJCknJ0elpaWqqqpqNG7p0qUaP368AoGAUlJSNHLkSC1btiyitRG2QNflS/LCt3X/9jljdKza1/bQQb31/3ZoT9VRSUa7K49q3zfH5fl14ewb6et9R/WHv2/WW6u36fW3t+rAkWr1TIjRgSPVev3trdpYtq9d67USuKFQSOnp6XLduntHu66rvn37KhQKNRuXkZERfhwMBrVnzx4bJQKIMidqfDlydKza06FjtfIb3FCy4c18j1d7WrF+l1w3oPhYV47jKD7WlesGtGzdznatKeru2pua2qujSwDQSbgByfP85i84koy+PdM1Ol7jKZgQI6fBfdVjXEf7D1crLS2p3eqxErjBYFB79+6V53lyXVee56m8vFzBYLDZuN27d2vw4MGSmp/xtkVl5WH5fie7NzKADuH5UkxM3YV8reedfOHbiHBUt+YbFxPQkeO1io91w0NO1Hjq0ytOFRWHzuiYrQW0lSWF1NRUZWVlqaSkRJJUUlKirKwspaSkNBo3ZswYLVy4UL7vq6qqSu+8846ys7NtlAggysTHBmRk1CPOVVKPGAUarCM0PCVLiHM1+rv95Xm+TtR4MsboRI0nz/M1ZtiAdq3JMcZYOR0sKyvTtGnTdPDgQfXu3VvFxcUaNGiQCgsLNWXKFF111VXyPE9z5szRBx98IEkqLCxUfn7+GR3nbM5w+cEZ0DV1xt9SaO0M11rg2nIuSwppaUlnfPnQFXWHPrtDjxJ9dkYdvqQAACBwAcAaAhcALCFwAcASAhcALCFwAcASAhcALIm6z1IINPxzkg7YvqvoDn12hx4l+uxKou4PHwCgs2JJAQAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsicrAffbZZ5WZmamtW7dKkj755BONHTtW2dnZmjRpkiorK8NjI/FapJ04cUJFRUUaPXq0cnNzNWPGDEnS9u3blZ+fr+zsbOXn52vHjh3hbSLxWqStWrVKeXl5GjdunHJzc7VixYqI9WKrz+LiYt14442N/v/ZET1Fut+W+ty/f78KCwuVnZ2t3Nxc3X///aqqqgpvE23v0xaZKPPZZ5+Ze+65x9xwww3m888/N77vm5EjR5r169cbY4yZP3++mTZtmjHGROQ1G+bOnWt+9atfGd/3jTHGVFRUGGOMmThxolm0aJExxphFixaZiRMnhreJxGuR5Pu+ue6668znn39ujDFm8+bN5pprrjGe53XpPtevX292795tfvSjH4V7i1TdHdlvS33u37/frF27Njxm3rx55uGHHzbGROa92NHv05ZEVeCeOHHC3H777Wbnzp3h/9CffvqpufXWW8NjKisrzTXXXGOMMRF5LdIOHz5shg4dag4fPtzo+X379pmhQ4ea2tpaY4wxtbW1ZujQoaaysjIir0Wa7/vme9/7nvnoo4+MMcZ8+OGHZvTo0VHTZ8Mgst2TzX6bTiwNLVu2zNx1113GmMi8FzvyfXoqUfVpYU899ZTGjh2rCy+8MPxcKBRSRkZG+HFKSop839eBAwci8lpycnJEe/zqq6+UnJysZ599VuvWrVNiYqJ+9rOfKSEhQenp6XJdV5Lkuq769u2rUCgkY0y7v5aSkhLRPh3H0e9+9zvdd9996tmzp44cOaIXX3xRoVAoqvqUZL2nju5Xknzf15tvvqkbb7wx/D2IpvfpqUTNGu6GDRu0adMmFRQUdHQpEVVbW6uvvvpK3/nOd/S3v/1NP//5z/XAAw/o6NGjHV1au6qtrdWLL76o5557TqtWrdLzzz+vqVOnRl2f3dXcuXPVs2dP3XnnnR1dilVRc4a7fv16bdu2TTfddJMkac+ePbrnnns0ceJE7d69OzyuqqpKjuMoOTlZwWCw3V+LtIyMDMXExCgnJ0eSdPXVV6tPnz5KSEjQ3r175XmeXNeV53kqLy9XMBiUMabdX4u0zZs3q7y8XEOHDpUkDR06VD169FB8fHxU9SlJwWDQak8d3W9xcbG+/PJLvfDCCwoEAuHvQTS9T08las5wJ0+erNWrV2vlypVauXKl+vXrp5dffln33nuvjh8/ro8++kiS9Oc//1k333yzJOnKK69s99ciLSUlRcOGDdMHH3wgqe6nzZWVlRo4cKCysrJUUlIiSSopKVFWVpZSUlKUmpra7q9FWr9+/bRnzx5t27ZNklRWVqZ9+/bpoosuiqo+JUWk7s7a75NPPqnPPvtM8+fPV1xcXPj5SLwXO/J9ekp2l4ztabhY//HHH5ucnBwzatQoc/fdd4d/qh+p1yJt586d5s477zQ5OTkmLy/PvPfee8YYY7744gtz2223mdGjR5vbbrvNlJWVhbeJxGuRtnjxYpOTk2Nyc3NNbm6uefvtt7t8n3PnzjUjRowwWVlZ5gc/+IG55ZZbOqSnSPfbUp9bt241l112mRk9erQZO3asGTt2rLnvvvvC20Tb+7Ql3PEBACyJmiUFAOjsCFwAsITABQBLCFwAsITABQBLCFx0a5mZmfryyy87ugx0EwQuAFhC4AJnoba2tqNLQBdE4CJqvPTSSxoxYoSGDBmi7OxsrVmzRhs3blR+fr6uu+46DR8+XHPmzFF1dXWL27/33nvKy8vTtddeqx/+8Id65plnwq/t2rVLmZmZWrhwoW644Qbdddddmjx5sv70pz812kdubq7eeeediPaJritqPrwG3du2bdv0+uuv669//avS09O1a9cu+b6vgwcP6uGHH9aVV16pPXv2qLCwUG+88YbuvvvuZvvo0aOHiouLdemll2rr1q2aNGmSsrKyNHLkyPCY9evXa+nSpQoEAlq5cqVeeeUVTZw4UZK0ZcsWlZeX63//939ttY0uhjNcRAXXdVVdXa2ysjLV1NSof//+GjBggK688kpdc801iomJUf/+/ZWfn6/169e3uI9hw4YpMzNTgUBAl19+uW699VZ9+OGHjcY88MAD6tmzpxISEjRy5Eh9+eWX4dvTLF68WDfffHOjD2UBGuIMF1Hhoosu0vTp0/XMM8/oiy++0PDhwzVt2jQdPXpU8+bN02effaZjx47J8zxdccUVLe7j008/1RNPPKH//ve/qqmpUXV1tcaMGdNoTL9+/cJfx8XFacyYMXrrrbd0//33q6SkRE8//XRE+0TXxhkuokZubq7efPNNrVq1So7j6IknntCsWbM0aNAgLV++XP/+9781depUnerzmh588EHddNNN+uc//6mPP/5Yd9xxR7OxjuM0evzjH/9YS5Ys0Zo1a9SjRw8NGTIkYv2h6yNwERW2bdumNWvWqLq6WnFxcYqPj5frujpy5IgSExOVmJiosrIyvfnmm6fcx5EjR3TeeecpPj5eGzduDH9ebGuGDBmiQCCgefPmaezYse3ZEqIQgYuoUF1drd/+9rcaNmyYhg8frqqqKk2dOlUPPfSQSkpKdO2112rGjBm65ZZbTrmPoqIiPf300xoyZIjmz5/f5g+rHjdunLZu3apx48a1VzuIUnweLnCOFi1apL/85S+tnj0DEme4wDk5duyY3njjDeXn53d0KegCCFzgLL3//vu6/vrrlZqaGr6pJ9AalhQAwBLOcAHAEgIXACwhcAHAEgIXACwhcAHAEgIXACz5/3eLRyZGQzLfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for col in emp_tsd.toPandas().columns:\n",
    "  try:\n",
    "    sns.lmplot(x= col, y='left', data=emp_tsd.toPandas())\n",
    "    plt.show()\n",
    "  except:\n",
    "    pass"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping these following columns since these do not seem to have any effect in Employee Left Status :\n",
    "'emp_title_id','birth_date','last_date','hire_date','emp_no','title_id','dept_no','first_name','last_name'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>sex</th>\n",
       "      <th>no_of_projects</th>\n",
       "      <th>last_performance_rating</th>\n",
       "      <th>left</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure_Years</th>\n",
       "      <th>title</th>\n",
       "      <th>salary</th>\n",
       "      <th>dept_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "      <td>60</td>\n",
       "      <td>8</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>60117</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>F</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>28</td>\n",
       "      <td>Staff</td>\n",
       "      <td>65828</td>\n",
       "      <td>Sales</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>54</td>\n",
       "      <td>27</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40006</td>\n",
       "      <td>Production</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>5</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>59</td>\n",
       "      <td>27</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40054</td>\n",
       "      <td>Production</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>M</td>\n",
       "      <td>6</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>58</td>\n",
       "      <td>24</td>\n",
       "      <td>Staff</td>\n",
       "      <td>78228</td>\n",
       "      <td>Human Resources</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>F</td>\n",
       "      <td>10</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>24</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40000</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>F</td>\n",
       "      <td>6</td>\n",
       "      <td>B</td>\n",
       "      <td>1</td>\n",
       "      <td>56</td>\n",
       "      <td>13</td>\n",
       "      <td>Staff</td>\n",
       "      <td>56724</td>\n",
       "      <td>Research</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>M</td>\n",
       "      <td>9</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>55</td>\n",
       "      <td>19</td>\n",
       "      <td>Assistant Engineer</td>\n",
       "      <td>46671</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>F</td>\n",
       "      <td>8</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>28</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>60929</td>\n",
       "      <td>Quality Management</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>24</td>\n",
       "      <td>Engineer</td>\n",
       "      <td>72488</td>\n",
       "      <td>Quality Management</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>F</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>23</td>\n",
       "      <td>Staff</td>\n",
       "      <td>42365</td>\n",
       "      <td>Customer Service</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>M</td>\n",
       "      <td>3</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>53</td>\n",
       "      <td>21</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>40000</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>M</td>\n",
       "      <td>6</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>28</td>\n",
       "      <td>Senior Staff</td>\n",
       "      <td>40000</td>\n",
       "      <td>Human Resources</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>M</td>\n",
       "      <td>9</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>57</td>\n",
       "      <td>26</td>\n",
       "      <td>Engineer</td>\n",
       "      <td>46168</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>M</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>54</td>\n",
       "      <td>26</td>\n",
       "      <td>Senior Staff</td>\n",
       "      <td>40000</td>\n",
       "      <td>Research</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>M</td>\n",
       "      <td>5</td>\n",
       "      <td>C</td>\n",
       "      <td>0</td>\n",
       "      <td>52</td>\n",
       "      <td>18</td>\n",
       "      <td>Staff</td>\n",
       "      <td>70889</td>\n",
       "      <td>Sales</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>F</td>\n",
       "      <td>10</td>\n",
       "      <td>A</td>\n",
       "      <td>0</td>\n",
       "      <td>55</td>\n",
       "      <td>20</td>\n",
       "      <td>Staff</td>\n",
       "      <td>71380</td>\n",
       "      <td>Marketing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>F</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>59</td>\n",
       "      <td>26</td>\n",
       "      <td>Senior Engineer</td>\n",
       "      <td>55881</td>\n",
       "      <td>development</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>M</td>\n",
       "      <td>4</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>14</td>\n",
       "      <td>Staff</td>\n",
       "      <td>44276</td>\n",
       "      <td>Research</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>M</td>\n",
       "      <td>7</td>\n",
       "      <td>B</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>22</td>\n",
       "      <td>Engineer</td>\n",
       "      <td>40000</td>\n",
       "      <td>Production</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sex  no_of_projects last_performance_rating  left  Age  Tenure_Years  \\\n",
       "0    M               9                       C     1   60             8   \n",
       "1    F               8                       B     0   49            28   \n",
       "2    M               1                       C     0   54            27   \n",
       "3    M               5                       A     0   59            27   \n",
       "4    M               6                       A     0   58            24   \n",
       "5    F              10                       B     0   60            24   \n",
       "6    F               6                       B     1   56            13   \n",
       "7    M               9                       C     0   55            19   \n",
       "8    F               8                       B     0   61            28   \n",
       "9    F               4                       A     0   50            24   \n",
       "10   F               6                       C     0   60            23   \n",
       "11   M               3                       A     0   53            21   \n",
       "12   M               6                       C     0   50            28   \n",
       "13   M               9                       B     0   57            26   \n",
       "14   M               4                       B     0   54            26   \n",
       "15   M               5                       C     0   52            18   \n",
       "16   F              10                       A     0   55            20   \n",
       "17   F               4                       B     0   59            26   \n",
       "18   M               4                       B     0   60            14   \n",
       "19   M               7                       B     0   61            22   \n",
       "\n",
       "                 title  salary           dept_name  \n",
       "0      Senior Engineer   60117         development  \n",
       "1                Staff   65828               Sales  \n",
       "2      Senior Engineer   40006          Production  \n",
       "3      Senior Engineer   40054          Production  \n",
       "4                Staff   78228     Human Resources  \n",
       "5      Senior Engineer   40000         development  \n",
       "6                Staff   56724            Research  \n",
       "7   Assistant Engineer   46671         development  \n",
       "8      Senior Engineer   60929  Quality Management  \n",
       "9             Engineer   72488  Quality Management  \n",
       "10               Staff   42365    Customer Service  \n",
       "11     Senior Engineer   40000         development  \n",
       "12        Senior Staff   40000     Human Resources  \n",
       "13            Engineer   46168         development  \n",
       "14        Senior Staff   40000            Research  \n",
       "15               Staff   70889               Sales  \n",
       "16               Staff   71380           Marketing  \n",
       "17     Senior Engineer   55881         development  \n",
       "18               Staff   44276            Research  \n",
       "19            Engineer   40000          Production  "
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = emp_tsd.drop('emp_title_id','birth_date','last_date','hire_date'\\\n",
    "                  ,'emp_no','title_id','dept_no','first_name','last_name')\n",
    "df.toPandas().head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "root\n",
      " |-- sex: string (nullable = true)\n",
      " |-- no_of_projects: integer (nullable = true)\n",
      " |-- last_performance_rating: string (nullable = true)\n",
      " |-- left: integer (nullable = true)\n",
      " |-- Age: integer (nullable = true)\n",
      " |-- Tenure_Years: integer (nullable = true)\n",
      " |-- title: string (nullable = true)\n",
      " |-- salary: integer (nullable = true)\n",
      " |-- dept_name: string (nullable = true)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df.printSchema()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Data Preparation for Modeling**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Label Encoding :"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### last_performance_rating :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-----------------------+\n",
      "|last_performance_rating|\n",
      "+-----------------------+\n",
      "|                      B|\n",
      "|                      C|\n",
      "|                      A|\n",
      "|                      S|\n",
      "|                    PIP|\n",
      "+-----------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df.select('last_performance_rating').distinct().show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Considering the the order as S > A > B > C > PIP, Encoding values as S = 5, A = 4, B = 3, C = 2, PIP = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.withColumn('last_performance_rating', F.when(df.last_performance_rating=='S',5)\\\n",
    "                                               .when(df.last_performance_rating=='A',4)\\\n",
    "                                               .when(df.last_performance_rating=='B',3)\\\n",
    "                                               .when(df.last_performance_rating=='C',2)\\\n",
    "                                               .otherwise(1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### title :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------------------+\n",
      "|             title|\n",
      "+------------------+\n",
      "|             Staff|\n",
      "|Assistant Engineer|\n",
      "|      Senior Staff|\n",
      "|   Senior Engineer|\n",
      "|          Engineer|\n",
      "|           Manager|\n",
      "|  Technique Leader|\n",
      "+------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df.select('title').distinct().show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Considering the the order as Staff < Senior Staff < Assistant Engineer < Engineer < Senior Engineer < Technique Leader < Manager, Encoding values as Manager = 7, Technique Leader = 6, Senior Engineer = 5, Engineer = 4, Assistant Engineer = 3, Senior Staff = 2, Staff = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.withColumn('title', F.when(df.title=='Manager',7)\\\n",
    "                             .when(df.title=='Technique Leader',6)\\\n",
    "                             .when(df.title=='Senior Engineer',5)\\\n",
    "                             .when(df.title=='Engineer',4)\\\n",
    "                             .when(df.title=='Assistant Engineer',3)\\\n",
    "                             .when(df.title=='Senior Staff',2)\\\n",
    "                             .otherwise(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+--------------+-----------------------+----+---+------------+-----+------+------------------+\n",
      "|sex|no_of_projects|last_performance_rating|left|Age|Tenure_Years|title|salary|         dept_name|\n",
      "+---+--------------+-----------------------+----+---+------------+-----+------+------------------+\n",
      "|  M|             9|                      2|   1| 60|           8|    5| 60117|       development|\n",
      "|  F|             8|                      3|   0| 49|          28|    1| 65828|             Sales|\n",
      "|  M|             1|                      2|   0| 54|          27|    5| 40006|        Production|\n",
      "|  M|             5|                      4|   0| 59|          27|    5| 40054|        Production|\n",
      "|  M|             6|                      4|   0| 58|          24|    1| 78228|   Human Resources|\n",
      "|  F|            10|                      3|   0| 60|          24|    5| 40000|       development|\n",
      "|  F|             6|                      3|   1| 56|          13|    1| 56724|          Research|\n",
      "|  M|             9|                      2|   0| 55|          19|    3| 46671|       development|\n",
      "|  F|             8|                      3|   0| 61|          28|    5| 60929|Quality Management|\n",
      "|  F|             4|                      4|   0| 50|          24|    4| 72488|Quality Management|\n",
      "|  F|             6|                      2|   0| 60|          23|    1| 42365|  Customer Service|\n",
      "|  M|             3|                      4|   0| 53|          21|    5| 40000|       development|\n",
      "|  M|             6|                      2|   0| 50|          28|    2| 40000|   Human Resources|\n",
      "|  M|             9|                      3|   0| 57|          26|    4| 46168|       development|\n",
      "|  M|             4|                      3|   0| 54|          26|    2| 40000|          Research|\n",
      "|  M|             5|                      2|   0| 52|          18|    1| 70889|             Sales|\n",
      "|  F|            10|                      4|   0| 55|          20|    1| 71380|         Marketing|\n",
      "|  F|             4|                      3|   0| 59|          26|    5| 55881|       development|\n",
      "|  M|             4|                      3|   0| 60|          14|    1| 44276|          Research|\n",
      "|  M|             7|                      3|   0| 61|          22|    4| 40000|        Production|\n",
      "+---+--------------+-----------------------+----+---+------------+-----+------+------------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Back_Up df after Label Encoding\n",
    "df_bkup = df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One-Hot-Encoding :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Encoding all categorical features\n",
    "from pyspark.ml.feature import StringIndexer, VectorAssembler, OneHotEncoderEstimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+---------+------------------+---------------+\n",
      "|sex|sex_Index|         dept_name|dept_name_Index|\n",
      "+---+---------+------------------+---------------+\n",
      "|  M|      0.0|       development|            0.0|\n",
      "|  F|      1.0|             Sales|            2.0|\n",
      "|  M|      0.0|        Production|            1.0|\n",
      "|  M|      0.0|        Production|            1.0|\n",
      "|  M|      0.0|   Human Resources|            6.0|\n",
      "|  F|      1.0|       development|            0.0|\n",
      "|  F|      1.0|          Research|            4.0|\n",
      "|  M|      0.0|       development|            0.0|\n",
      "|  F|      1.0|Quality Management|            5.0|\n",
      "|  F|      1.0|Quality Management|            5.0|\n",
      "+---+---------+------------------+---------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# create object of StringIndexer class and specify input and output column\n",
    "SI_sex = StringIndexer(inputCol='sex',outputCol='sex_Index')\n",
    "SI_dept_name = StringIndexer(inputCol='dept_name',outputCol='dept_name_Index')\n",
    "\n",
    "# transform the data\n",
    "df = SI_sex.fit(df).transform(df)\n",
    "df = SI_dept_name.fit(df).transform(df)\n",
    "\n",
    "# view the transformed data\n",
    "df.select('sex', 'sex_Index', 'dept_name', 'dept_name_Index').show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+---------+-------------+------------------+---------------+-------------+\n",
      "|sex|sex_Index|      sex_vec|         dept_name|dept_name_Index|dept_name_vec|\n",
      "+---+---------+-------------+------------------+---------------+-------------+\n",
      "|  M|      0.0|(1,[0],[1.0])|       development|            0.0|(8,[0],[1.0])|\n",
      "|  F|      1.0|    (1,[],[])|             Sales|            2.0|(8,[2],[1.0])|\n",
      "|  M|      0.0|(1,[0],[1.0])|        Production|            1.0|(8,[1],[1.0])|\n",
      "|  M|      0.0|(1,[0],[1.0])|        Production|            1.0|(8,[1],[1.0])|\n",
      "|  M|      0.0|(1,[0],[1.0])|   Human Resources|            6.0|(8,[6],[1.0])|\n",
      "|  F|      1.0|    (1,[],[])|       development|            0.0|(8,[0],[1.0])|\n",
      "|  F|      1.0|    (1,[],[])|          Research|            4.0|(8,[4],[1.0])|\n",
      "|  M|      0.0|(1,[0],[1.0])|       development|            0.0|(8,[0],[1.0])|\n",
      "|  F|      1.0|    (1,[],[])|Quality Management|            5.0|(8,[5],[1.0])|\n",
      "|  F|      1.0|    (1,[],[])|Quality Management|            5.0|(8,[5],[1.0])|\n",
      "+---+---------+-------------+------------------+---------------+-------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# create object and specify input and output column\n",
    "OHE_sex = OneHotEncoderEstimator(inputCols=['sex_Index'],outputCols=['sex_vec'])\n",
    "OHE_dept_name = OneHotEncoderEstimator(inputCols=['dept_name_Index'],outputCols=['dept_name_vec'])\n",
    "\n",
    "# transform the data\n",
    "df = OHE_sex.fit(df).transform(df)\n",
    "df = OHE_dept_name.fit(df).transform(df)\n",
    "\n",
    "# view and transform the data\n",
    "df.select('sex', 'sex_Index','sex_vec', \\\n",
    "          'dept_name', 'dept_name_Index','dept_name_vec').show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.withColumnRenamed('left', 'label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>sex</th>\n",
       "      <th>no_of_projects</th>\n",
       "      <th>last_performance_rating</th>\n",
       "      <th>label</th>\n",
       "      <th>Age</th>\n",
       "      <th>Tenure_Years</th>\n",
       "      <th>title</th>\n",
       "      <th>salary</th>\n",
       "      <th>dept_name</th>\n",
       "      <th>sex_Index</th>\n",
       "      <th>dept_name_Index</th>\n",
       "      <th>sex_vec</th>\n",
       "      <th>dept_name_vec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>60</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>60117</td>\n",
       "      <td>development</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>(1.0)</td>\n",
       "      <td>(1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>F</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>65828</td>\n",
       "      <td>Sales</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>(0.0)</td>\n",
       "      <td>(0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>54</td>\n",
       "      <td>27</td>\n",
       "      <td>5</td>\n",
       "      <td>40006</td>\n",
       "      <td>Production</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>(1.0)</td>\n",
       "      <td>(0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>59</td>\n",
       "      <td>27</td>\n",
       "      <td>5</td>\n",
       "      <td>40054</td>\n",
       "      <td>Production</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>(1.0)</td>\n",
       "      <td>(0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>M</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>58</td>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "      <td>78228</td>\n",
       "      <td>Human Resources</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>(1.0)</td>\n",
       "      <td>(0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  sex  no_of_projects  last_performance_rating  label  Age  Tenure_Years  \\\n",
       "0   M               9                        2      1   60             8   \n",
       "1   F               8                        3      0   49            28   \n",
       "2   M               1                        2      0   54            27   \n",
       "3   M               5                        4      0   59            27   \n",
       "4   M               6                        4      0   58            24   \n",
       "\n",
       "   title  salary        dept_name  sex_Index  dept_name_Index sex_vec  \\\n",
       "0      5   60117      development        0.0              0.0   (1.0)   \n",
       "1      1   65828            Sales        1.0              2.0   (0.0)   \n",
       "2      5   40006       Production        0.0              1.0   (1.0)   \n",
       "3      5   40054       Production        0.0              1.0   (1.0)   \n",
       "4      1   78228  Human Resources        0.0              6.0   (1.0)   \n",
       "\n",
       "                              dept_name_vec  \n",
       "0  (1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)  \n",
       "1  (0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0)  \n",
       "2  (0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)  \n",
       "3  (0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)  \n",
       "4  (0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0)  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.toPandas().head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Assembler :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['sex',\n",
       " 'no_of_projects',\n",
       " 'last_performance_rating',\n",
       " 'label',\n",
       " 'Age',\n",
       " 'Tenure_Years',\n",
       " 'title',\n",
       " 'salary',\n",
       " 'dept_name',\n",
       " 'sex_Index',\n",
       " 'dept_name_Index',\n",
       " 'sex_vec',\n",
       " 'dept_name_vec']"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "assembler = VectorAssembler(inputCols = [\n",
    " 'no_of_projects',\n",
    " 'last_performance_rating',\n",
    " 'Age',\n",
    " 'Tenure_Years',\n",
    " 'title',\n",
    " 'salary',\n",
    " 'sex_vec',\n",
    " 'dept_name_vec',], outputCol = \"features\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-----+---------------------------------------------------------------+\n",
      "|label|features                                                       |\n",
      "+-----+---------------------------------------------------------------+\n",
      "|1    |(15,[0,1,2,3,4,5,6,7],[9.0,2.0,60.0,8.0,5.0,60117.0,1.0,1.0])  |\n",
      "|0    |(15,[0,1,2,3,4,5,9],[8.0,3.0,49.0,28.0,1.0,65828.0,1.0])       |\n",
      "|0    |(15,[0,1,2,3,4,5,6,8],[1.0,2.0,54.0,27.0,5.0,40006.0,1.0,1.0]) |\n",
      "|0    |(15,[0,1,2,3,4,5,6,8],[5.0,4.0,59.0,27.0,5.0,40054.0,1.0,1.0]) |\n",
      "|0    |(15,[0,1,2,3,4,5,6,13],[6.0,4.0,58.0,24.0,1.0,78228.0,1.0,1.0])|\n",
      "|0    |(15,[0,1,2,3,4,5,7],[10.0,3.0,60.0,24.0,5.0,40000.0,1.0])      |\n",
      "|1    |(15,[0,1,2,3,4,5,11],[6.0,3.0,56.0,13.0,1.0,56724.0,1.0])      |\n",
      "|0    |(15,[0,1,2,3,4,5,6,7],[9.0,2.0,55.0,19.0,3.0,46671.0,1.0,1.0]) |\n",
      "|0    |(15,[0,1,2,3,4,5,12],[8.0,3.0,61.0,28.0,5.0,60929.0,1.0])      |\n",
      "|0    |(15,[0,1,2,3,4,5,12],[4.0,4.0,50.0,24.0,4.0,72488.0,1.0])      |\n",
      "+-----+---------------------------------------------------------------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "data = assembler.transform(df).select('label','features')\n",
    "data.show(10, truncate=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Train Test Split**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df, test_df = data.randomSplit( [0.7, 0.3], seed = 42 )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **Model Building**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyspark.ml import Pipeline\n",
    "from pyspark.ml.classification import RandomForestClassifier\n",
    "from pyspark.ml.feature import *\n",
    "from pyspark.ml.evaluation import MulticlassClassificationEvaluator, BinaryClassificationEvaluator\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train Function to build the model and check accuacy\n",
    "def train(train, test, classifier):\n",
    "    print(classifier)\n",
    "    model = classifier.fit(train)\n",
    "\n",
    "    pred = model.transform(test)\n",
    "\n",
    "    eval_accuracy = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"accuracy\"))\n",
    "\n",
    "    eval_precision = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"weightedPrecision\"))\n",
    "\n",
    "    eval_recall = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"weightedRecall\"))\n",
    "\n",
    "    eval_f1 = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"f1\"))\n",
    "\n",
    "    accuracy = eval_accuracy.evaluate(pred)\n",
    "\n",
    "    precision =  eval_precision.evaluate(pred)\n",
    "\n",
    "    recall =  eval_recall.evaluate(pred)\n",
    "\n",
    "    f1 =  eval_f1.evaluate(pred)\n",
    "\n",
    "    print(f\"\"\"\n",
    "    Accuracy  = {accuracy}\n",
    "    Error     = {1-accuracy}\n",
    "    Precision = {precision}\n",
    "    Recall    = {recall}\n",
    "    F1        = {f1}\"\"\")\n",
    "\n",
    "    return model, pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest Classifier Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc = RandomForestClassifier(featuresCol=\"features\",\n",
    "                              labelCol=\"label\",\n",
    "                              numTrees=50,\n",
    "                              maxDepth=5,\n",
    "                              featureSubsetStrategy='onethird')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier_ce110dc17493\n",
      "\n",
      "    Accuracy  = 0.9980555834925744\n",
      "    Error     = 0.001944416507425606\n",
      "    Precision = 0.9980594815941188\n",
      "    Recall    = 0.9980555834925743\n",
      "    F1        = 0.9980471469145747\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "|label|            features|       rawPrediction|         probability|prediction|\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3410460390212...|[0.98682092078042...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4408732192368...|[0.98881746438473...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4410653136910...|[0.98882130627382...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3607091886313...|[0.98721418377262...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4445703932617...|[0.98889140786523...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.6444542333341...|[0.99288908466668...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.5585614521556...|[0.99117122904311...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.6448511827771...|[0.99289702365554...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.5646074606149...|[0.99129214921229...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4159737882359...|[0.98831947576471...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4412544641398...|[0.98882508928279...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3588172907594...|[0.98717634581518...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3572341205677...|[0.98714468241135...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3572341205677...|[0.98714468241135...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3597101852755...|[0.98719420370551...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3612933554672...|[0.98722586710934...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3595180908213...|[0.98719036181642...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.4454157490611...|[0.98890831498122...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3589339239854...|[0.98717867847970...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[49.3591260184395...|[0.98718252036879...|       0.0|\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_rf, pred_rf = train(train_df, test_df, rfc)\n",
    "pred_rf.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Area under ROC (Random Forest Model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under ROC on train dataset 0.9999458862996691\n",
      "Area under ROC on test dataset 0.9999469372959502\n"
     ]
    }
   ],
   "source": [
    "print('Area under ROC on train dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(model_rf.transform(train_df)))\n",
    "print('Area under ROC on test dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(pred_rf))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Logistic Regression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyspark.ml.classification import LogisticRegression\n",
    "mlr = LogisticRegression(maxIter=10, \n",
    "                         regParam=0.3, \n",
    "                         elasticNetParam=0.8, \n",
    "                         family=\"multinomial\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression_b9d5d0551431\n",
      "\n",
      "    Accuracy  = 0.9004101154357617\n",
      "    Error     = 0.09958988456423834\n",
      "    Precision = 0.8107383759790415\n",
      "    Recall    = 0.9004101154357617\n",
      "    F1        = 0.8532246480840692\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "|label|            features|       rawPrediction|         probability|prediction|\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10238003882660...|[0.90067615376621...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10018650552051...|[0.90028300239394...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10091768328921...|[0.90041420637067...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10347680547964...|[0.90087221183491...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10128327217355...|[0.90047975076500...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10164886105790...|[0.90054525678490...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10274562771095...|[0.90074154476001...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10347680547964...|[0.90087221183491...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10238003882660...|[0.90067615376621...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10347680547964...|[0.90087221183491...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.1031112165953,...|[0.90080689744418...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.1031112165953,...|[0.90080689744418...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.1031112165953,...|[0.90080689744418...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10347680547964...|[0.90087221183491...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10091768328921...|[0.90041420637067...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10201444994225...|[0.90061072444658...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10274562771095...|[0.90074154476001...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.09908973886746...|[0.90008590821584...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10274562771095...|[0.90074154476001...|       0.0|\n",
      "|    0|(15,[0,1,2,3,4,5]...|[1.10091768328921...|[0.90041420637067...|       0.0|\n",
      "+-----+--------------------+--------------------+--------------------+----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_lr, pred_lr = train(train_df, test_df, mlr)\n",
    "pred_lr.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Area under ROC (Logistic Regression Model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under ROC on train dataset 0.9999504281834609\n",
      "Area under ROC on test dataset 0.999949764261694\n"
     ]
    }
   ],
   "source": [
    "print('Area under ROC on train dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(model_lr.transform(train_df)))\n",
    "print('Area under ROC on test dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(pred_lr))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Saving the Models to HDFS :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_rf.write().overwrite().save(\"random_forest.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_lr.write().overwrite().save(\"logistic_regression.model\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Pipeline**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pipeline Stages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---+--------------+-----------------------+-----+---+------------+-----+------+------------------+\n",
      "|sex|no_of_projects|last_performance_rating|label|Age|Tenure_Years|title|salary|         dept_name|\n",
      "+---+--------------+-----------------------+-----+---+------------+-----+------+------------------+\n",
      "|  M|             9|                      2|    1| 60|           8|    5| 60117|       development|\n",
      "|  F|             8|                      3|    0| 49|          28|    1| 65828|             Sales|\n",
      "|  M|             1|                      2|    0| 54|          27|    5| 40006|        Production|\n",
      "|  M|             5|                      4|    0| 59|          27|    5| 40054|        Production|\n",
      "|  M|             6|                      4|    0| 58|          24|    1| 78228|   Human Resources|\n",
      "|  F|            10|                      3|    0| 60|          24|    5| 40000|       development|\n",
      "|  F|             6|                      3|    1| 56|          13|    1| 56724|          Research|\n",
      "|  M|             9|                      2|    0| 55|          19|    3| 46671|       development|\n",
      "|  F|             8|                      3|    0| 61|          28|    5| 60929|Quality Management|\n",
      "|  F|             4|                      4|    0| 50|          24|    4| 72488|Quality Management|\n",
      "|  F|             6|                      2|    0| 60|          23|    1| 42365|  Customer Service|\n",
      "|  M|             3|                      4|    0| 53|          21|    5| 40000|       development|\n",
      "|  M|             6|                      2|    0| 50|          28|    2| 40000|   Human Resources|\n",
      "|  M|             9|                      3|    0| 57|          26|    4| 46168|       development|\n",
      "|  M|             4|                      3|    0| 54|          26|    2| 40000|          Research|\n",
      "|  M|             5|                      2|    0| 52|          18|    1| 70889|             Sales|\n",
      "|  F|            10|                      4|    0| 55|          20|    1| 71380|         Marketing|\n",
      "|  F|             4|                      3|    0| 59|          26|    5| 55881|       development|\n",
      "|  M|             4|                      3|    0| 60|          14|    1| 44276|          Research|\n",
      "|  M|             7|                      3|    0| 61|          22|    4| 40000|        Production|\n",
      "+---+--------------+-----------------------+-----+---+------------+-----+------+------------------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Preparing database  \n",
    "df = employees.join(titles, employees.emp_title_id == titles.title_id)\\\n",
    "    .join(salaries, on='emp_no')\\\n",
    "    .join(dept_emp, on='emp_no')\\\n",
    "    .join(departments, on='dept_no')\\\n",
    "    .withColumnRenamed('left2', 'label')\\\n",
    "    .drop('emp_title_id','birth_date','last_date','hire_date'\\\n",
    "                  ,'emp_no','title_id','dept_no','first_name','last_name')\n",
    "# Label Encoding\n",
    "df_pl = df.withColumn('last_performance_rating', F.when(df.last_performance_rating=='S',5)\\\n",
    "                                               .when(df.last_performance_rating=='A',4)\\\n",
    "                                               .when(df.last_performance_rating=='B',3)\\\n",
    "                                               .when(df.last_performance_rating=='C',2)\\\n",
    "                                               .otherwise(1))\\\n",
    "          .withColumn('title', F.when(df.title=='Manager',7)\\\n",
    "                             .when(df.title=='Technique Leader',6)\\\n",
    "                             .when(df.title=='Senior Engineer',5)\\\n",
    "                             .when(df.title=='Engineer',4)\\\n",
    "                             .when(df.title=='Assistant Engineer',3)\\\n",
    "                             .when(df.title=='Senior Staff',2)\\\n",
    "                             .otherwise(1))\n",
    "df_pl.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "conCols = [\n",
    "'no_of_projects',\n",
    " 'last_performance_rating',\n",
    " 'Age',\n",
    " 'Tenure_Years',\n",
    " 'title',\n",
    " 'salary']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "catCols = ['sex', 'dept_name']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "# String Indexer\n",
    "indexers = [StringIndexer(inputCol=column, outputCol=column+\"_Index\") for column in catCols ]\n",
    "\n",
    "# One Hot Encoder Estimator\n",
    "encoders = OneHotEncoderEstimator(inputCols=[i.getOutputCol() for i in indexers], \\\n",
    "                                  outputCols=[i.getOutputCol()+\"_vec\" for i in indexers])\n",
    "\n",
    "# Vector Assembler\n",
    "assembler = VectorAssembler(inputCols = encoders.getOutputCols() + conCols, outputCol = \"features\")\n",
    "\n",
    "# ML Models\n",
    "rfc = RandomForestClassifier(featuresCol=\"features\",\n",
    "                              labelCol=\"label\",\n",
    "                              numTrees=50,\n",
    "                              maxDepth=5,\n",
    "                              featureSubsetStrategy='onethird')\n",
    "mlr = LogisticRegression(maxIter=10, \n",
    "                         regParam=0.3, \n",
    "                         elasticNetParam=0.8, \n",
    "                         family=\"multinomial\")\n",
    "# Creating Pipelines\n",
    "pipeline_rfc = Pipeline(stages = indexers + [encoders, assembler, rfc])\n",
    "pipeline_mlr = Pipeline(stages = indexers + [encoders, assembler, mlr])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = df_pl.randomSplit( [0.7, 0.3], seed = 42 )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Checking Accuracy, Error, Precision, Recall, F1 Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracyCheck(pred):\n",
    "    eval_accuracy = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"accuracy\"))\n",
    "\n",
    "    eval_precision = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"weightedPrecision\"))\n",
    "\n",
    "    eval_recall = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"weightedRecall\"))\n",
    "\n",
    "    eval_f1 = (MulticlassClassificationEvaluator\n",
    "        (labelCol=\"label\", predictionCol=\"prediction\", metricName=\"f1\"))\n",
    "\n",
    "    accuracy = eval_accuracy.evaluate(pred)\n",
    "\n",
    "    precision =  eval_precision.evaluate(pred)\n",
    "\n",
    "    recall =  eval_recall.evaluate(pred)\n",
    "\n",
    "    f1 =  eval_f1.evaluate(pred)\n",
    "\n",
    "    print(f\"\"\"\n",
    "    Accuracy  = {accuracy}\n",
    "    Error     = {1-accuracy}\n",
    "    Precision = {precision}\n",
    "    Recall    = {recall}\n",
    "    F1        = {f1}\"\"\")\n",
    "\n",
    "    return pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest Classifier Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    Accuracy  = 0.9976756400371004\n",
      "    Error     = 0.0023243599628995915\n",
      "    Precision = 0.9976805986159328\n",
      "    Recall    = 0.9976756400371003\n",
      "    F1        = 0.9976638430104859\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "|label|features                                                     |prediction|\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "|0    |(15,[1,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,24.0,5.0,50527.0])|0.0       |\n",
      "|0    |(15,[1,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,25.0,4.0,40000.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,27.0,4.0,58610.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,21.0,3.0,40000.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,23.0,1.0,47408.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,25.0,4.0,40000.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,25.0,5.0,40000.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,26.0,1.0,40000.0])|0.0       |\n",
      "|0    |(15,[3,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,27.0,1.0,70833.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,27.0,4.0,40000.0])|0.0       |\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_rfc = pipeline_rfc.fit(train)\n",
    "pred_rfc = model_rfc.transform(test)\n",
    "pred_rfc = accuracyCheck(pred_rfc)\n",
    "pred_rfc.select('label', 'features', 'prediction').show(10, truncate=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Area under ROC (Random Forest Model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under ROC on train dataset 0.9999525956543478\n",
      "Area under ROC on test dataset 0.9999157516113358\n"
     ]
    }
   ],
   "source": [
    "print('Area under ROC on train dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(model_rfc.transform(train)))\n",
    "print('Area under ROC on test dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(pred_rfc))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Logistic Regression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    Accuracy  = 0.8998960742901203\n",
      "    Error     = 0.1001039257098797\n",
      "    Precision = 0.8098129445227698\n",
      "    Recall    = 0.8998960742901203\n",
      "    F1        = 0.8524813072476602\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "|label|features                                                     |prediction|\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "|0    |(15,[1,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,24.0,5.0,50527.0])|0.0       |\n",
      "|0    |(15,[1,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,25.0,4.0,40000.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,48.0,27.0,4.0,58610.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,21.0,3.0,40000.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,23.0,1.0,47408.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,25.0,4.0,40000.0])|0.0       |\n",
      "|0    |(15,[2,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,25.0,5.0,40000.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,26.0,1.0,40000.0])|0.0       |\n",
      "|0    |(15,[3,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,27.0,1.0,70833.0])|0.0       |\n",
      "|0    |(15,[4,9,10,11,12,13,14],[1.0,1.0,1.0,49.0,27.0,4.0,40000.0])|0.0       |\n",
      "+-----+-------------------------------------------------------------+----------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_mlr = pipeline_mlr.fit(train)\n",
    "pred_mlr = model_mlr.transform(test)\n",
    "pred_mlr = accuracyCheck(pred_mlr)\n",
    "pred_mlr.select('label', 'features', 'prediction').show(10, truncate=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Area under ROC (Logistic Regression Model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under ROC on train dataset 0.9999554346690951\n",
      "Area under ROC on test dataset 0.9999374877649229\n"
     ]
    }
   ],
   "source": [
    "print('Area under ROC on train dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(model_mlr.transform(train)))\n",
    "print('Area under ROC on test dataset',BinaryClassificationEvaluator(labelCol=\"label\", \\\n",
    "        rawPredictionCol=\"rawPrediction\", metricName=\"areaUnderROC\").evaluate(pred_mlr))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Accuracies between the built models and the Pipeline models are very close. The reason behind the slight change in the accuracies is that the eariler case, the train & test split was performed after fitting the assembler but in case of ML pipeline, the assembler is inside the stages so assembler is fitting on split datasets seperately as a part of the pipeline. This is also clearly visible in the features column as well. So, this was a good test of the pipeline models in terms of accuracy and we can conclude that the ML Pipeline is working properly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### -End-"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}