eHealth Take-Home.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.\n",
    "**Assume each customer type is equally common. A customer walks into your store and purchases a product. What's the probability that this customer is of Type A, given that they purchased a product?**\n",
    "\n",
    "    Set-up:\n",
    "\n",
    "    Pr(purchase/Type A)=0.01\n",
    "\n",
    "    Pr(purchase/type B)=0.1\n",
    "\n",
    "    ________________________\n",
    "\n",
    "    Pr(Type A)=Pr(Type B)=0.5\n",
    "\n",
    "    Pr(purchase)=Pr(purchase/Type A)*Pr(Type A)+Pr(purchase/Type B)*Pr(Type B)\n",
    "\n",
    "    =(0.01*0.5) + (0.1*0.5)\n",
    "    =0.055"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.\n",
    "**This is how the probability found in question 1 would change if more customers were of Type A:**\n",
    "    \n",
    "\n",
    "    Pr(Type A/purchase)=Pr(purchase/Type A)*Pr(Type A)/Pr(purchase)\n",
    "\n",
    "    =(0.01*0.5)/0.055\n",
    "    =0.091\n",
    "    \n",
    "    This is an increase to Pr(purchase). So the answer to the question would be:\n",
    "    \n",
    "    I. **Increase**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {},
   "outputs": [],
   "source": [
    "def A_Prob(TypeA):\n",
    "    #P(purchase/Type A)*P(Type A) + P(purchase/Type B)*P(Type B)\n",
    "    \n",
    "    TypeB = 1-TypeA\n",
    "    \n",
    "    TypeAProb = (0.01*TypeA) + (.1*TypeB)\n",
    "    \n",
    "    print(TypeAProb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.055\n"
     ]
    }
   ],
   "source": [
    "A_Prob(.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data exploration idea I would like to pursue:\n",
    "\n",
    "# How does health insurance type (attributes/cost) differ across states and over time?\n",
    "\n",
    "\n",
    "I. My exploration idea is to see what sort of insurance plans cost most and if these vary across different regions of America and years. This interests me because there is a huge variety of health plans tailored to how healthy people are/if they have prior conditions, how concerned people are about preventative health. So I was curious about the different price points for each and how these may vary from year-to-year and state-to-state.\n",
    "\n",
    "Answers to this would give insight into different demographic make-up of each state which would influence health insurance sellers' target groups. Health insurance types' consumption over time would give insight into how a certain plans may grow or lose popularity.\n",
    "\n",
    "\n",
    "II. Scope of what data is required:\n",
    "### Data tables to look at in order to answer this exploration idea:\n",
    "    -BenefitsCostSharing\n",
    "    -Rate\n",
    "    \n",
    "III. Methodology I want to pursue:\n",
    "    a. Clean up the data (deal with any NaN's and Outliers)\n",
    "    b. Present descriptive Statistics (location of data (mean/median), spread of data (how variable is it around mean/median)\n",
    "    c. Present comparative data across time\n",
    "    d. Visual map of US showing differing types of healthcare plans across states\n",
    "    \n",
    "IV. Outstanding concerns with analysis plan:\n",
    "    a. Data is filled with a lot of NaN's\n",
    "    b. Data input is not standardized. For example, to represent $0, it is represented as $0 or 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script><script>requirejs.config({paths: { 'plotly': ['https://cdn.plot.ly/plotly-latest.min']},});if(!window._Plotly) {require(['plotly'],function(plotly) {window._Plotly=plotly;});}</script>"
      ],
      "text/vnd.plotly.v1+html": [
       "<script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script><script>requirejs.config({paths: { 'plotly': ['https://cdn.plot.ly/plotly-latest.min']},});if(!window._Plotly) {require(['plotly'],function(plotly) {window._Plotly=plotly;});}</script>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import plotly.plotly as py\n",
    "import plotly.graph_objs as go\n",
    "from plotly.offline import download_plotlyjs,init_notebook_mode,plot,iplot\n",
    "init_notebook_mode(connected=True)\n",
    "import seaborn as sns \n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random as rnd\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "import matplotlib.gridspec as gridspec\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from numpy import genfromtxt\n",
    "from scipy import stats\n",
    "from scipy.stats import multivariate_normal\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.metrics import recall_score , average_precision_score\n",
    "from sklearn.metrics import precision_score, precision_recall_curve\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Exploring Health Plans Across Time and States"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## a. Data Integrity/Clean Data\n",
    "\n",
    "a. Need to inspect data and see if there are any NaN's that need to be taken out/replaced with 0's.\n",
    "b. If there are outliers, take these out as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/IPython/core/interactiveshell.py:2728: DtypeWarning:\n",
      "\n",
      "Columns (3,6,9,10,16,17,18) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "\n"
     ]
    },
    {
     "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>BenefitName</th>\n",
       "      <th>BusinessYear</th>\n",
       "      <th>CoinsInnTier1</th>\n",
       "      <th>CoinsInnTier2</th>\n",
       "      <th>CoinsOutofNet</th>\n",
       "      <th>CopayInnTier1</th>\n",
       "      <th>CopayInnTier2</th>\n",
       "      <th>CopayOutofNet</th>\n",
       "      <th>EHBVarReason</th>\n",
       "      <th>Exclusions</th>\n",
       "      <th>...</th>\n",
       "      <th>LimitUnit</th>\n",
       "      <th>MinimumStay</th>\n",
       "      <th>PlanId</th>\n",
       "      <th>QuantLimitOnSvc</th>\n",
       "      <th>RowNumber</th>\n",
       "      <th>SourceName</th>\n",
       "      <th>StandardComponentId</th>\n",
       "      <th>StateCode</th>\n",
       "      <th>StateCode2</th>\n",
       "      <th>VersionNum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>2014</td>\n",
       "      <td>20%</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20%</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>Above EHB</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>Dollars per Year</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21989AK0010001-00</td>\n",
       "      <td>Yes</td>\n",
       "      <td>68</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>21989AK0010001</td>\n",
       "      <td>AK</td>\n",
       "      <td>AK</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Dental Check-Up for Children</td>\n",
       "      <td>2014</td>\n",
       "      <td>20%</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20%</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>Substantially Equal</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>Visit(s) per 6 Months</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21989AK0010001-00</td>\n",
       "      <td>Yes</td>\n",
       "      <td>104</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>21989AK0010001</td>\n",
       "      <td>AK</td>\n",
       "      <td>AK</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>2014</td>\n",
       "      <td>40%</td>\n",
       "      <td>NaN</td>\n",
       "      <td>40%</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>Substantially Equal</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21989AK0010001-00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>110</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>21989AK0010001</td>\n",
       "      <td>AK</td>\n",
       "      <td>AK</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>2014</td>\n",
       "      <td>50%</td>\n",
       "      <td>NaN</td>\n",
       "      <td>50%</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>Additional EHB Benefit</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21989AK0010001-00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>111</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>21989AK0010001</td>\n",
       "      <td>AK</td>\n",
       "      <td>AK</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Major Dental Care - Child</td>\n",
       "      <td>2014</td>\n",
       "      <td>50%</td>\n",
       "      <td>NaN</td>\n",
       "      <td>50%</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No Charge</td>\n",
       "      <td>Substantially Equal</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21989AK0010001-00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>112</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>21989AK0010001</td>\n",
       "      <td>AK</td>\n",
       "      <td>AK</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       BenefitName  BusinessYear CoinsInnTier1 CoinsInnTier2  \\\n",
       "0  Routine Dental Services (Adult)          2014           20%           NaN   \n",
       "1     Dental Check-Up for Children          2014           20%           NaN   \n",
       "2        Basic Dental Care - Child          2014           40%           NaN   \n",
       "3              Orthodontia - Child          2014           50%           NaN   \n",
       "4        Major Dental Care - Child          2014           50%           NaN   \n",
       "\n",
       "  CoinsOutofNet CopayInnTier1 CopayInnTier2 CopayOutofNet  \\\n",
       "0           20%     No Charge           NaN     No Charge   \n",
       "1           20%     No Charge           NaN     No Charge   \n",
       "2           40%     No Charge           NaN     No Charge   \n",
       "3           50%     No Charge           NaN     No Charge   \n",
       "4           50%     No Charge           NaN     No Charge   \n",
       "\n",
       "             EHBVarReason Exclusions     ...                  LimitUnit  \\\n",
       "0               Above EHB        NaN     ...           Dollars per Year   \n",
       "1     Substantially Equal        NaN     ...      Visit(s) per 6 Months   \n",
       "2     Substantially Equal        NaN     ...                        NaN   \n",
       "3  Additional EHB Benefit        NaN     ...                        NaN   \n",
       "4     Substantially Equal        NaN     ...                        NaN   \n",
       "\n",
       "  MinimumStay             PlanId QuantLimitOnSvc RowNumber SourceName  \\\n",
       "0         NaN  21989AK0010001-00             Yes        68       HIOS   \n",
       "1         NaN  21989AK0010001-00             Yes       104       HIOS   \n",
       "2         NaN  21989AK0010001-00             NaN       110       HIOS   \n",
       "3         NaN  21989AK0010001-00             NaN       111       HIOS   \n",
       "4         NaN  21989AK0010001-00             NaN       112       HIOS   \n",
       "\n",
       "  StandardComponentId StateCode StateCode2  VersionNum  \n",
       "0      21989AK0010001        AK         AK           6  \n",
       "1      21989AK0010001        AK         AK           6  \n",
       "2      21989AK0010001        AK         AK           6  \n",
       "3      21989AK0010001        AK         AK           6  \n",
       "4      21989AK0010001        AK         AK           6  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 284,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BenCosts = pd.read_csv(\"BenefitsCostSharing.csv\")\n",
    "BenCosts.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (b. Comparative Statistics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "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>BusinessYear</th>\n",
       "      <th>IssuerId</th>\n",
       "      <th>IssuerId2</th>\n",
       "      <th>LimitQty</th>\n",
       "      <th>MinimumStay</th>\n",
       "      <th>RowNumber</th>\n",
       "      <th>VersionNum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5.048408e+06</td>\n",
       "      <td>5.048408e+06</td>\n",
       "      <td>5.048408e+06</td>\n",
       "      <td>687683.000000</td>\n",
       "      <td>16727.000000</td>\n",
       "      <td>5.048408e+06</td>\n",
       "      <td>5.048408e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.015127e+03</td>\n",
       "      <td>5.220360e+04</td>\n",
       "      <td>5.220360e+04</td>\n",
       "      <td>142.399374</td>\n",
       "      <td>47.937048</td>\n",
       "      <td>9.839018e+01</td>\n",
       "      <td>7.637094e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>7.563664e-01</td>\n",
       "      <td>2.592498e+04</td>\n",
       "      <td>2.592498e+04</td>\n",
       "      <td>1723.372817</td>\n",
       "      <td>5.892297</td>\n",
       "      <td>2.177890e+01</td>\n",
       "      <td>3.803627e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2.014000e+03</td>\n",
       "      <td>1.004600e+04</td>\n",
       "      <td>1.004600e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.100000e+01</td>\n",
       "      <td>1.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.015000e+03</td>\n",
       "      <td>3.253600e+04</td>\n",
       "      <td>3.253600e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>8.000000e+01</td>\n",
       "      <td>5.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.015000e+03</td>\n",
       "      <td>4.839600e+04</td>\n",
       "      <td>4.839600e+04</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>9.900000e+01</td>\n",
       "      <td>7.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.016000e+03</td>\n",
       "      <td>7.498000e+04</td>\n",
       "      <td>7.498000e+04</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>1.170000e+02</td>\n",
       "      <td>9.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.016000e+03</td>\n",
       "      <td>9.996900e+04</td>\n",
       "      <td>9.996900e+04</td>\n",
       "      <td>75000.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>1.560000e+02</td>\n",
       "      <td>2.400000e+01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       BusinessYear      IssuerId     IssuerId2       LimitQty   MinimumStay  \\\n",
       "count  5.048408e+06  5.048408e+06  5.048408e+06  687683.000000  16727.000000   \n",
       "mean   2.015127e+03  5.220360e+04  5.220360e+04     142.399374     47.937048   \n",
       "std    7.563664e-01  2.592498e+04  2.592498e+04    1723.372817      5.892297   \n",
       "min    2.014000e+03  1.004600e+04  1.004600e+04       1.000000      5.000000   \n",
       "25%    2.015000e+03  3.253600e+04  3.253600e+04       1.000000     48.000000   \n",
       "50%    2.015000e+03  4.839600e+04  4.839600e+04      20.000000     48.000000   \n",
       "75%    2.016000e+03  7.498000e+04  7.498000e+04      40.000000     48.000000   \n",
       "max    2.016000e+03  9.996900e+04  9.996900e+04   75000.000000     90.000000   \n",
       "\n",
       "          RowNumber    VersionNum  \n",
       "count  5.048408e+06  5.048408e+06  \n",
       "mean   9.839018e+01  7.637094e+00  \n",
       "std    2.177890e+01  3.803627e+00  \n",
       "min    6.100000e+01  1.000000e+00  \n",
       "25%    8.000000e+01  5.000000e+00  \n",
       "50%    9.900000e+01  7.000000e+00  \n",
       "75%    1.170000e+02  9.000000e+00  \n",
       "max    1.560000e+02  2.400000e+01  "
      ]
     },
     "execution_count": 285,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BenCosts.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5048408 entries, 0 to 5048407\n",
      "Data columns (total 32 columns):\n",
      "BenefitName            object\n",
      "BusinessYear           int64\n",
      "CoinsInnTier1          object\n",
      "CoinsInnTier2          object\n",
      "CoinsOutofNet          object\n",
      "CopayInnTier1          object\n",
      "CopayInnTier2          object\n",
      "CopayOutofNet          object\n",
      "EHBVarReason           object\n",
      "Exclusions             object\n",
      "Explanation            object\n",
      "ImportDate             object\n",
      "IsCovered              object\n",
      "IsEHB                  object\n",
      "IsExclFromInnMOOP      object\n",
      "IsExclFromOonMOOP      object\n",
      "IsStateMandate         object\n",
      "IsSubjToDedTier1       object\n",
      "IsSubjToDedTier2       object\n",
      "IssuerId               int64\n",
      "IssuerId2              int64\n",
      "LimitQty               float64\n",
      "LimitUnit              object\n",
      "MinimumStay            float64\n",
      "PlanId                 object\n",
      "QuantLimitOnSvc        object\n",
      "RowNumber              int64\n",
      "SourceName             object\n",
      "StandardComponentId    object\n",
      "StateCode              object\n",
      "StateCode2             object\n",
      "VersionNum             int64\n",
      "dtypes: float64(2), int64(5), object(25)\n",
      "memory usage: 1.2+ GB\n"
     ]
    }
   ],
   "source": [
    "BenCosts.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BenefitName                  0\n",
       "BusinessYear                 0\n",
       "CoinsInnTier1          1113847\n",
       "CoinsInnTier2          4571587\n",
       "CoinsOutofNet          1113847\n",
       "CopayInnTier1          1113847\n",
       "CopayInnTier2          4571587\n",
       "CopayOutofNet          1113849\n",
       "EHBVarReason           3020737\n",
       "Exclusions             4572247\n",
       "Explanation            4075700\n",
       "ImportDate                   0\n",
       "IsCovered               215980\n",
       "IsEHB                  1817362\n",
       "IsExclFromInnMOOP       983986\n",
       "IsExclFromOonMOOP       982626\n",
       "IsStateMandate         4250463\n",
       "IsSubjToDedTier1       2465877\n",
       "IsSubjToDedTier2       2466054\n",
       "IssuerId                     0\n",
       "IssuerId2                    0\n",
       "LimitQty               4360725\n",
       "LimitUnit              4360539\n",
       "MinimumStay            5031681\n",
       "PlanId                       0\n",
       "QuantLimitOnSvc        3264532\n",
       "RowNumber                    0\n",
       "SourceName                   0\n",
       "StandardComponentId          0\n",
       "StateCode                    0\n",
       "StateCode2                   0\n",
       "VersionNum                   0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 287,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BenCosts.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Due to huge amount of NaN's will not eliminate rows that contain them, but rather replace NaN's with 0's to not compromise data's wealth of information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [],
   "source": [
    "CleanBenCosts = BenCosts.fillna(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check to see if NaN's are left."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BenefitName            0\n",
       "BusinessYear           0\n",
       "CoinsInnTier1          0\n",
       "CoinsInnTier2          0\n",
       "CoinsOutofNet          0\n",
       "CopayInnTier1          0\n",
       "CopayInnTier2          0\n",
       "CopayOutofNet          0\n",
       "EHBVarReason           0\n",
       "Exclusions             0\n",
       "Explanation            0\n",
       "ImportDate             0\n",
       "IsCovered              0\n",
       "IsEHB                  0\n",
       "IsExclFromInnMOOP      0\n",
       "IsExclFromOonMOOP      0\n",
       "IsStateMandate         0\n",
       "IsSubjToDedTier1       0\n",
       "IsSubjToDedTier2       0\n",
       "IssuerId               0\n",
       "IssuerId2              0\n",
       "LimitQty               0\n",
       "LimitUnit              0\n",
       "MinimumStay            0\n",
       "PlanId                 0\n",
       "QuantLimitOnSvc        0\n",
       "RowNumber              0\n",
       "SourceName             0\n",
       "StandardComponentId    0\n",
       "StateCode              0\n",
       "StateCode2             0\n",
       "VersionNum             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 289,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CleanBenCosts.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that data is clean, let us explore top benefit plans across states."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## c. Benefit Plans Across Time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"4\" halign=\"left\">BenefitName</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>unique</th>\n",
       "      <th>top</th>\n",
       "      <th>freq</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>1164869</td>\n",
       "      <td>496</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>18719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>2079286</td>\n",
       "      <td>517</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>31269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>1804253</td>\n",
       "      <td>429</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>27389</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             BenefitName                                         \n",
       "                   count unique                        top   freq\n",
       "BusinessYear                                                     \n",
       "2014             1164869    496  Basic Dental Care - Child  18719\n",
       "2015             2079286    517        Orthodontia - Adult  31269\n",
       "2016             1804253    429        Orthodontia - Adult  27389"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CleanBenCosts[[\"BusinessYear\",\"BenefitName\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For 2014: 1.2 million plans bought, close to 500 different ones with adult dental care toping the list at close to 19,000 bought.\n",
    "\n",
    "For 2015: 2 million plans bought, over 500 different ones with adult orthodontic dental care toping the list with over 30,000 of these plans bought. 2015 was a huge revenue boost for health insurance companies/brokers.\n",
    "    \n",
    "For 2016: 1.8 million plans bought, around 430 different ones only and adult orthodontic dental care toping the list with close to 27,000 of these plans bought.    \n",
    "    \n",
    "Another cut of the data that we can do is naming the most popular benefit plans per state. This could give way to state demographics, which was not supplied in available dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let us see what the top benefit plans are spread across states. We just need to substitute out BusinessYear with StateCode feature."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## d. Benefit Plans Across States"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"4\" halign=\"left\">BenefitName</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>unique</th>\n",
       "      <th>top</th>\n",
       "      <th>freq</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>StateCode</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AK</th>\n",
       "      <td>41320</td>\n",
       "      <td>84</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AL</th>\n",
       "      <td>28417</td>\n",
       "      <td>115</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AR</th>\n",
       "      <td>68064</td>\n",
       "      <td>93</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>1077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AZ</th>\n",
       "      <td>221612</td>\n",
       "      <td>195</td>\n",
       "      <td>Dental Check-Up for Children</td>\n",
       "      <td>3345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE</th>\n",
       "      <td>31370</td>\n",
       "      <td>110</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FL</th>\n",
       "      <td>364742</td>\n",
       "      <td>194</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>5130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GA</th>\n",
       "      <td>196579</td>\n",
       "      <td>151</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>2893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HI</th>\n",
       "      <td>6741</td>\n",
       "      <td>76</td>\n",
       "      <td>Major Dental Care - Child</td>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IA</th>\n",
       "      <td>110221</td>\n",
       "      <td>104</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>1727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <td>27313</td>\n",
       "      <td>71</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IL</th>\n",
       "      <td>320533</td>\n",
       "      <td>157</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>4299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IN</th>\n",
       "      <td>148531</td>\n",
       "      <td>109</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>2347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KS</th>\n",
       "      <td>64279</td>\n",
       "      <td>82</td>\n",
       "      <td>Accidental Dental</td>\n",
       "      <td>1087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LA</th>\n",
       "      <td>91816</td>\n",
       "      <td>125</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>1491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ME</th>\n",
       "      <td>61834</td>\n",
       "      <td>99</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MI</th>\n",
       "      <td>189225</td>\n",
       "      <td>205</td>\n",
       "      <td>Major Dental Care - Child</td>\n",
       "      <td>3248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MO</th>\n",
       "      <td>95152</td>\n",
       "      <td>127</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>1577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MS</th>\n",
       "      <td>46895</td>\n",
       "      <td>104</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MT</th>\n",
       "      <td>58132</td>\n",
       "      <td>75</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>1049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NC</th>\n",
       "      <td>94290</td>\n",
       "      <td>98</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>1537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ND</th>\n",
       "      <td>48269</td>\n",
       "      <td>77</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NE</th>\n",
       "      <td>66935</td>\n",
       "      <td>90</td>\n",
       "      <td>Dental Check-Up for Children</td>\n",
       "      <td>1238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NH</th>\n",
       "      <td>44305</td>\n",
       "      <td>98</td>\n",
       "      <td>Major Dental Care - Child</td>\n",
       "      <td>707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NJ</th>\n",
       "      <td>86107</td>\n",
       "      <td>100</td>\n",
       "      <td>Major Dental Care - Child</td>\n",
       "      <td>1357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NM</th>\n",
       "      <td>56669</td>\n",
       "      <td>78</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NV</th>\n",
       "      <td>62743</td>\n",
       "      <td>121</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OH</th>\n",
       "      <td>331045</td>\n",
       "      <td>193</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>5092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OK</th>\n",
       "      <td>133639</td>\n",
       "      <td>83</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>2136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OR</th>\n",
       "      <td>93502</td>\n",
       "      <td>123</td>\n",
       "      <td>Orthodontia - Adult</td>\n",
       "      <td>1344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PA</th>\n",
       "      <td>260866</td>\n",
       "      <td>157</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>4208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SC</th>\n",
       "      <td>111907</td>\n",
       "      <td>85</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>1817</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SD</th>\n",
       "      <td>47730</td>\n",
       "      <td>79</td>\n",
       "      <td>Basic Dental Care - Adult</td>\n",
       "      <td>843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TN</th>\n",
       "      <td>138154</td>\n",
       "      <td>129</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>2073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TX</th>\n",
       "      <td>465164</td>\n",
       "      <td>180</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>6441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>UT</th>\n",
       "      <td>98421</td>\n",
       "      <td>101</td>\n",
       "      <td>Routine Dental Services (Adult)</td>\n",
       "      <td>1527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VA</th>\n",
       "      <td>142236</td>\n",
       "      <td>145</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>2256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WI</th>\n",
       "      <td>512587</td>\n",
       "      <td>146</td>\n",
       "      <td>Dental Check-Up for Children</td>\n",
       "      <td>7135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WV</th>\n",
       "      <td>32638</td>\n",
       "      <td>77</td>\n",
       "      <td>Orthodontia - Child</td>\n",
       "      <td>625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WY</th>\n",
       "      <td>48425</td>\n",
       "      <td>92</td>\n",
       "      <td>Basic Dental Care - Child</td>\n",
       "      <td>807</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          BenefitName                                              \n",
       "                count unique                              top  freq\n",
       "StateCode                                                          \n",
       "AK              41320     84        Basic Dental Care - Adult   720\n",
       "AL              28417    115              Orthodontia - Child   653\n",
       "AR              68064     93              Orthodontia - Adult  1077\n",
       "AZ             221612    195     Dental Check-Up for Children  3345\n",
       "DE              31370    110              Orthodontia - Child   602\n",
       "FL             364742    194  Routine Dental Services (Adult)  5130\n",
       "GA             196579    151        Basic Dental Care - Adult  2893\n",
       "HI               6741     76        Major Dental Care - Child   110\n",
       "IA             110221    104        Basic Dental Care - Child  1727\n",
       "ID              27313     71              Orthodontia - Child   447\n",
       "IL             320533    157              Orthodontia - Adult  4299\n",
       "IN             148531    109              Orthodontia - Adult  2347\n",
       "KS              64279     82                Accidental Dental  1087\n",
       "LA              91816    125              Orthodontia - Child  1491\n",
       "ME              61834     99        Basic Dental Care - Adult   975\n",
       "MI             189225    205        Major Dental Care - Child  3248\n",
       "MO              95152    127        Basic Dental Care - Child  1577\n",
       "MS              46895    104        Basic Dental Care - Child   725\n",
       "MT              58132     75        Basic Dental Care - Adult  1049\n",
       "NC              94290     98  Routine Dental Services (Adult)  1537\n",
       "ND              48269     77        Basic Dental Care - Adult   974\n",
       "NE              66935     90     Dental Check-Up for Children  1238\n",
       "NH              44305     98        Major Dental Care - Child   707\n",
       "NJ              86107    100        Major Dental Care - Child  1357\n",
       "NM              56669     78        Basic Dental Care - Adult   828\n",
       "NV              62743    121  Routine Dental Services (Adult)   930\n",
       "OH             331045    193        Basic Dental Care - Child  5092\n",
       "OK             133639     83              Orthodontia - Child  2136\n",
       "OR              93502    123              Orthodontia - Adult  1344\n",
       "PA             260866    157  Routine Dental Services (Adult)  4208\n",
       "SC             111907     85        Basic Dental Care - Adult  1817\n",
       "SD              47730     79        Basic Dental Care - Adult   843\n",
       "TN             138154    129  Routine Dental Services (Adult)  2073\n",
       "TX             465164    180        Basic Dental Care - Child  6441\n",
       "UT              98421    101  Routine Dental Services (Adult)  1527\n",
       "VA             142236    145        Basic Dental Care - Child  2256\n",
       "WI             512587    146     Dental Check-Up for Children  7135\n",
       "WV              32638     77              Orthodontia - Child   625\n",
       "WY              48425     92        Basic Dental Care - Child   807"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "CleanBenCosts[[\"StateCode\",\"BenefitName\"]].groupby('StateCode').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So interesting! Dental care is at the top of each state accounted for. I would not have thought this since Dental Insurance is not an essential health benefit for those over 18 years old, but for children it is."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let us look visually at how many benefit plans vary across states."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [],
   "source": [
    "CleanBenCosts[[\"StateCode\",\"BenefitName\"]].groupby('StateCode').count().sort_values(\"BenefitName\")\n",
    "Distinct_State = CleanBenCosts.StateCode.unique()\n",
    "Benefit = []\n",
    "\n",
    "for state in Distinct_State:\n",
    "    state_benefit =  len(CleanBenCosts[CleanBenCosts[\"StateCode\"] == state])    \n",
    "    Benefit.append(state_benefit)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is best to have bar counts in descending order from top to bottom in order to visualize where most plans are sold in the US. This would give suggestion to where to target marketing teams/where to target sales efforts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame(\n",
    "    {'state': Distinct_State,\n",
    "     'Count' : Benefit\n",
    "     })\n",
    "\n",
    "df = df.sort_values(\"Count\", ascending=False).reset_index(drop=True)\n",
    "\n",
    "f, ax = plt.subplots(figsize=(15, 15)) \n",
    "ax.set_yticklabels(df.state, rotation='horizontal', fontsize='large')\n",
    "g = sns.barplot(y = df.state, x= df.Count)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Great! We can see from here that number of benefit plans across states only reach beyond 300,000 mark for WI, TX, FL, OH and IL. It is surprising that WI tops the list since TX is the second largest state in the US and FL is the state with the most old people."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To take this visualization further, we can look at benefit plans on a US map. Gradients of color represent volume of benefit plans per state. Moving cursor over a state will give you the exact number of benefit plans along with the two-letter state abbreviation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "linkText": "Export to plot.ly",
        "plotlyServerURL": "https://plot.ly",
        "showLink": false
       },
       "data": [
        {
         "colorbar": {
          "title": {
           "text": "No of Benefit plans"
          }
         },
         "colorscale": "Cividis",
         "locationmode": "USA-states",
         "locations": [
          "WI",
          "TX",
          "FL",
          "OH",
          "IL",
          "PA",
          "AZ",
          "GA",
          "MI",
          "IN",
          "VA",
          "TN",
          "OK",
          "SC",
          "IA",
          "UT",
          "MO",
          "NC",
          "OR",
          "LA",
          "NJ",
          "AR",
          "NE",
          "KS",
          "NV",
          "ME",
          "MT",
          "NM",
          "WY",
          "ND",
          "SD",
          "MS",
          "NH",
          "AK",
          "WV",
          "DE",
          "AL",
          "ID",
          "HI"
         ],
         "marker": {
          "line": {
           "color": "rgb(255,255,255)",
           "width": 2
          }
         },
         "text": [
          "WI",
          "TX",
          "FL",
          "OH",
          "IL",
          "PA",
          "AZ",
          "GA",
          "MI",
          "IN",
          "VA",
          "TN",
          "OK",
          "SC",
          "IA",
          "UT",
          "MO",
          "NC",
          "OR",
          "LA",
          "NJ",
          "AR",
          "NE",
          "KS",
          "NV",
          "ME",
          "MT",
          "NM",
          "WY",
          "ND",
          "SD",
          "MS",
          "NH",
          "AK",
          "WV",
          "DE",
          "AL",
          "ID",
          "HI"
         ],
         "type": "choropleth",
         "uid": "ab07dd1d-81ed-4506-8495-cd6c41902bbb",
         "z": [
          512587,
          465164,
          364742,
          331045,
          320533,
          260866,
          221612,
          196579,
          189225,
          148531,
          142236,
          138154,
          133639,
          111907,
          110221,
          98421,
          95152,
          94290,
          93502,
          91816,
          86107,
          68064,
          66935,
          64279,
          62743,
          61834,
          58132,
          56669,
          48425,
          48269,
          47730,
          46895,
          44305,
          41320,
          32638,
          31370,
          28417,
          27313,
          6741
         ]
        }
       ],
       "layout": {
        "geo": {
         "lakecolor": "rgb(255,255,255)",
         "scope": "usa",
         "showlakes": true
        },
        "title": {
         "text": "2014-2016 Benefit Plans by State (Hover for Breakdown)"
        }
       }
      },
      "text/html": [
       "<div id=\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\" style=\"height: 525px; width: 100%;\" class=\"plotly-graph-div\"></div><script type=\"text/javascript\">require([\"plotly\"], function(Plotly) { window.PLOTLYENV=window.PLOTLYENV || {};window.PLOTLYENV.BASE_URL=\"https://plot.ly\";Plotly.newPlot(\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\", [{\"colorbar\": {\"title\": {\"text\": \"No of Benefit plans\"}}, \"colorscale\": \"Cividis\", \"locationmode\": \"USA-states\", \"locations\": [\"WI\", \"TX\", \"FL\", \"OH\", \"IL\", \"PA\", \"AZ\", \"GA\", \"MI\", \"IN\", \"VA\", \"TN\", \"OK\", \"SC\", \"IA\", \"UT\", \"MO\", \"NC\", \"OR\", \"LA\", \"NJ\", \"AR\", \"NE\", \"KS\", \"NV\", \"ME\", \"MT\", \"NM\", \"WY\", \"ND\", \"SD\", \"MS\", \"NH\", \"AK\", \"WV\", \"DE\", \"AL\", \"ID\", \"HI\"], \"marker\": {\"line\": {\"color\": \"rgb(255,255,255)\", \"width\": 2}}, \"text\": [\"WI\", \"TX\", \"FL\", \"OH\", \"IL\", \"PA\", \"AZ\", \"GA\", \"MI\", \"IN\", \"VA\", \"TN\", \"OK\", \"SC\", \"IA\", \"UT\", \"MO\", \"NC\", \"OR\", \"LA\", \"NJ\", \"AR\", \"NE\", \"KS\", \"NV\", \"ME\", \"MT\", \"NM\", \"WY\", \"ND\", \"SD\", \"MS\", \"NH\", \"AK\", \"WV\", \"DE\", \"AL\", \"ID\", \"HI\"], \"z\": [512587, 465164, 364742, 331045, 320533, 260866, 221612, 196579, 189225, 148531, 142236, 138154, 133639, 111907, 110221, 98421, 95152, 94290, 93502, 91816, 86107, 68064, 66935, 64279, 62743, 61834, 58132, 56669, 48425, 48269, 47730, 46895, 44305, 41320, 32638, 31370, 28417, 27313, 6741], \"type\": \"choropleth\", \"uid\": \"ab07dd1d-81ed-4506-8495-cd6c41902bbb\"}], {\"geo\": {\"lakecolor\": \"rgb(255,255,255)\", \"scope\": \"usa\", \"showlakes\": true}, \"title\": {\"text\": \"2014-2016 Benefit Plans by State (Hover for Breakdown)\"}}, {\"showLink\": false, \"linkText\": \"Export to plot.ly\", \"plotlyServerURL\": \"https://plot.ly\"})});</script><script type=\"text/javascript\">window.addEventListener(\"resize\", function(){window._Plotly.Plots.resize(document.getElementById(\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\"));});</script>"
      ],
      "text/vnd.plotly.v1+html": [
       "<div id=\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\" style=\"height: 525px; width: 100%;\" class=\"plotly-graph-div\"></div><script type=\"text/javascript\">require([\"plotly\"], function(Plotly) { window.PLOTLYENV=window.PLOTLYENV || {};window.PLOTLYENV.BASE_URL=\"https://plot.ly\";Plotly.newPlot(\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\", [{\"colorbar\": {\"title\": {\"text\": \"No of Benefit plans\"}}, \"colorscale\": \"Cividis\", \"locationmode\": \"USA-states\", \"locations\": [\"WI\", \"TX\", \"FL\", \"OH\", \"IL\", \"PA\", \"AZ\", \"GA\", \"MI\", \"IN\", \"VA\", \"TN\", \"OK\", \"SC\", \"IA\", \"UT\", \"MO\", \"NC\", \"OR\", \"LA\", \"NJ\", \"AR\", \"NE\", \"KS\", \"NV\", \"ME\", \"MT\", \"NM\", \"WY\", \"ND\", \"SD\", \"MS\", \"NH\", \"AK\", \"WV\", \"DE\", \"AL\", \"ID\", \"HI\"], \"marker\": {\"line\": {\"color\": \"rgb(255,255,255)\", \"width\": 2}}, \"text\": [\"WI\", \"TX\", \"FL\", \"OH\", \"IL\", \"PA\", \"AZ\", \"GA\", \"MI\", \"IN\", \"VA\", \"TN\", \"OK\", \"SC\", \"IA\", \"UT\", \"MO\", \"NC\", \"OR\", \"LA\", \"NJ\", \"AR\", \"NE\", \"KS\", \"NV\", \"ME\", \"MT\", \"NM\", \"WY\", \"ND\", \"SD\", \"MS\", \"NH\", \"AK\", \"WV\", \"DE\", \"AL\", \"ID\", \"HI\"], \"z\": [512587, 465164, 364742, 331045, 320533, 260866, 221612, 196579, 189225, 148531, 142236, 138154, 133639, 111907, 110221, 98421, 95152, 94290, 93502, 91816, 86107, 68064, 66935, 64279, 62743, 61834, 58132, 56669, 48425, 48269, 47730, 46895, 44305, 41320, 32638, 31370, 28417, 27313, 6741], \"type\": \"choropleth\", \"uid\": \"ab07dd1d-81ed-4506-8495-cd6c41902bbb\"}], {\"geo\": {\"lakecolor\": \"rgb(255,255,255)\", \"scope\": \"usa\", \"showlakes\": true}, \"title\": {\"text\": \"2014-2016 Benefit Plans by State (Hover for Breakdown)\"}}, {\"showLink\": false, \"linkText\": \"Export to plot.ly\", \"plotlyServerURL\": \"https://plot.ly\"})});</script><script type=\"text/javascript\">window.addEventListener(\"resize\", function(){window._Plotly.Plots.resize(document.getElementById(\"0ba4c867-f4cc-43c1-88ce-ba9092c0c340\"));});</script>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data = dict(type = 'choropleth',\n",
    "           locations = df['state'],\n",
    "           locationmode = 'USA-states',\n",
    "           colorscale = 'Cividis',\n",
    "            text = df['state'],\n",
    "            marker = dict (line = dict(color = 'rgb(255,255,255)',width=2)),\n",
    "           z = df['Count'],\n",
    "           colorbar = {'title':'No of Benefit plans'})\n",
    "\n",
    "layout = dict(title = '2014-2016 Benefit Plans by State (Hover for Breakdown)',\n",
    "         geo=dict(scope = 'usa',showlakes = True,lakecolor='rgb(255,255,255)')) \n",
    "\n",
    "choromap2 = go.Figure(data = [data],layout=layout)\n",
    "iplot(choromap2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With this map, can see that some states do not have data at all, such as CA, WA, CO, MN, KY, MD, NY and states left of NH. This is a data integrity issue since CA and NY are huge states right beside TX. A lot of data on the US as a whole is missing then. This data hole impacts analysis conclusions more so if there were final comments made in relation to US as a whole."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some census data would help since the density of plans across states do not account for everyone in each state. Some data on how many people are in each state would help with seeing how many uninsured could be targeted for health insurance sales. Other Census data that would help would be to see what are health benefits cut across gender, socioeconmic status, etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Now, let us explore cost of healthcare costs across time and states."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now move forward to a different dataset called Rate. Here, we will see how different insurance for individuals who do not use tobacco and those who do vary across time and across states.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "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>BusinessYear</th>\n",
       "      <th>StateCode</th>\n",
       "      <th>IssuerId</th>\n",
       "      <th>SourceName</th>\n",
       "      <th>VersionNum</th>\n",
       "      <th>ImportDate</th>\n",
       "      <th>IssuerId2</th>\n",
       "      <th>FederalTIN</th>\n",
       "      <th>RateEffectiveDate</th>\n",
       "      <th>RateExpirationDate</th>\n",
       "      <th>...</th>\n",
       "      <th>IndividualRate</th>\n",
       "      <th>IndividualTobaccoRate</th>\n",
       "      <th>Couple</th>\n",
       "      <th>PrimarySubscriberAndOneDependent</th>\n",
       "      <th>PrimarySubscriberAndTwoDependents</th>\n",
       "      <th>PrimarySubscriberAndThreeOrMoreDependents</th>\n",
       "      <th>CoupleAndOneDependent</th>\n",
       "      <th>CoupleAndTwoDependents</th>\n",
       "      <th>CoupleAndThreeOrMoreDependents</th>\n",
       "      <th>RowNumber</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2014</td>\n",
       "      <td>AK</td>\n",
       "      <td>21989</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>6</td>\n",
       "      <td>2014-03-19 07:06:49</td>\n",
       "      <td>21989</td>\n",
       "      <td>93-0438772</td>\n",
       "      <td>2014-01-01</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>...</td>\n",
       "      <td>29.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2014</td>\n",
       "      <td>AK</td>\n",
       "      <td>21989</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>6</td>\n",
       "      <td>2014-03-19 07:06:49</td>\n",
       "      <td>21989</td>\n",
       "      <td>93-0438772</td>\n",
       "      <td>2014-01-01</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>...</td>\n",
       "      <td>36.95</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73.9</td>\n",
       "      <td>107.61</td>\n",
       "      <td>107.61</td>\n",
       "      <td>107.61</td>\n",
       "      <td>144.56</td>\n",
       "      <td>144.56</td>\n",
       "      <td>144.56</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2014</td>\n",
       "      <td>AK</td>\n",
       "      <td>21989</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>6</td>\n",
       "      <td>2014-03-19 07:06:49</td>\n",
       "      <td>21989</td>\n",
       "      <td>93-0438772</td>\n",
       "      <td>2014-01-01</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>...</td>\n",
       "      <td>36.95</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73.9</td>\n",
       "      <td>107.61</td>\n",
       "      <td>107.61</td>\n",
       "      <td>107.61</td>\n",
       "      <td>144.56</td>\n",
       "      <td>144.56</td>\n",
       "      <td>144.56</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2014</td>\n",
       "      <td>AK</td>\n",
       "      <td>21989</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>6</td>\n",
       "      <td>2014-03-19 07:06:49</td>\n",
       "      <td>21989</td>\n",
       "      <td>93-0438772</td>\n",
       "      <td>2014-01-01</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>...</td>\n",
       "      <td>32.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2014</td>\n",
       "      <td>AK</td>\n",
       "      <td>21989</td>\n",
       "      <td>HIOS</td>\n",
       "      <td>6</td>\n",
       "      <td>2014-03-19 07:06:49</td>\n",
       "      <td>21989</td>\n",
       "      <td>93-0438772</td>\n",
       "      <td>2014-01-01</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>...</td>\n",
       "      <td>32.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   BusinessYear StateCode  IssuerId SourceName  VersionNum  \\\n",
       "0          2014        AK     21989       HIOS           6   \n",
       "1          2014        AK     21989       HIOS           6   \n",
       "2          2014        AK     21989       HIOS           6   \n",
       "3          2014        AK     21989       HIOS           6   \n",
       "4          2014        AK     21989       HIOS           6   \n",
       "\n",
       "            ImportDate  IssuerId2  FederalTIN RateEffectiveDate  \\\n",
       "0  2014-03-19 07:06:49      21989  93-0438772        2014-01-01   \n",
       "1  2014-03-19 07:06:49      21989  93-0438772        2014-01-01   \n",
       "2  2014-03-19 07:06:49      21989  93-0438772        2014-01-01   \n",
       "3  2014-03-19 07:06:49      21989  93-0438772        2014-01-01   \n",
       "4  2014-03-19 07:06:49      21989  93-0438772        2014-01-01   \n",
       "\n",
       "  RateExpirationDate    ...     IndividualRate IndividualTobaccoRate Couple  \\\n",
       "0         2014-12-31    ...              29.00                   NaN    NaN   \n",
       "1         2014-12-31    ...              36.95                   NaN   73.9   \n",
       "2         2014-12-31    ...              36.95                   NaN   73.9   \n",
       "3         2014-12-31    ...              32.00                   NaN    NaN   \n",
       "4         2014-12-31    ...              32.00                   NaN    NaN   \n",
       "\n",
       "  PrimarySubscriberAndOneDependent  PrimarySubscriberAndTwoDependents  \\\n",
       "0                              NaN                                NaN   \n",
       "1                           107.61                             107.61   \n",
       "2                           107.61                             107.61   \n",
       "3                              NaN                                NaN   \n",
       "4                              NaN                                NaN   \n",
       "\n",
       "   PrimarySubscriberAndThreeOrMoreDependents  CoupleAndOneDependent  \\\n",
       "0                                        NaN                    NaN   \n",
       "1                                     107.61                 144.56   \n",
       "2                                     107.61                 144.56   \n",
       "3                                        NaN                    NaN   \n",
       "4                                        NaN                    NaN   \n",
       "\n",
       "   CoupleAndTwoDependents  CoupleAndThreeOrMoreDependents  RowNumber  \n",
       "0                     NaN                             NaN         14  \n",
       "1                  144.56                          144.56         14  \n",
       "2                  144.56                          144.56         15  \n",
       "3                     NaN                             NaN         15  \n",
       "4                     NaN                             NaN         16  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 295,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate = pd.read_csv(\"Rate.csv\")\n",
    "Rate.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on the data table above, there is opportunity to explore plans that were offered (no information regarding who bought these plans). We can look at year they were offered, states they offered in, how long these plans lasted, and numerous different rates per plan for different individuals and different type of family plans. However, we will focus on certain columns that will answer our exploration idea: BusinessYear, StateCode and relevant rate columns. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## a. Data Integrity/Clean Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us first take a look at how many NaN's there are and in the dataset and what is the data type (int/str) of each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 0, 0, 7762096)"
      ]
     },
     "execution_count": 296,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate.BusinessYear.isnull().sum(), Rate.StateCode.isnull().sum(), Rate.IndividualRate.isnull().sum(), Rate.IndividualTobaccoRate.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12653504"
      ]
     },
     "execution_count": 297,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate.Couple.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 12694445 entries, 0 to 12694444\n",
      "Data columns (total 24 columns):\n",
      "BusinessYear                                 int64\n",
      "StateCode                                    object\n",
      "IssuerId                                     int64\n",
      "SourceName                                   object\n",
      "VersionNum                                   int64\n",
      "ImportDate                                   object\n",
      "IssuerId2                                    int64\n",
      "FederalTIN                                   object\n",
      "RateEffectiveDate                            object\n",
      "RateExpirationDate                           object\n",
      "PlanId                                       object\n",
      "RatingAreaId                                 object\n",
      "Tobacco                                      object\n",
      "Age                                          object\n",
      "IndividualRate                               float64\n",
      "IndividualTobaccoRate                        float64\n",
      "Couple                                       float64\n",
      "PrimarySubscriberAndOneDependent             float64\n",
      "PrimarySubscriberAndTwoDependents            float64\n",
      "PrimarySubscriberAndThreeOrMoreDependents    float64\n",
      "CoupleAndOneDependent                        float64\n",
      "CoupleAndTwoDependents                       float64\n",
      "CoupleAndThreeOrMoreDependents               float64\n",
      "RowNumber                                    int64\n",
      "dtypes: float64(9), int64(5), object(10)\n",
      "memory usage: 2.3+ GB\n"
     ]
    }
   ],
   "source": [
    "Rate.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (b. Comparative Statistics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "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>BusinessYear</th>\n",
       "      <th>IssuerId</th>\n",
       "      <th>VersionNum</th>\n",
       "      <th>IssuerId2</th>\n",
       "      <th>IndividualRate</th>\n",
       "      <th>IndividualTobaccoRate</th>\n",
       "      <th>Couple</th>\n",
       "      <th>PrimarySubscriberAndOneDependent</th>\n",
       "      <th>PrimarySubscriberAndTwoDependents</th>\n",
       "      <th>PrimarySubscriberAndThreeOrMoreDependents</th>\n",
       "      <th>CoupleAndOneDependent</th>\n",
       "      <th>CoupleAndTwoDependents</th>\n",
       "      <th>CoupleAndThreeOrMoreDependents</th>\n",
       "      <th>RowNumber</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.269444e+07</td>\n",
       "      <td>1.269444e+07</td>\n",
       "      <td>1.269444e+07</td>\n",
       "      <td>1.269444e+07</td>\n",
       "      <td>1.269444e+07</td>\n",
       "      <td>4.932349e+06</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>40941.000000</td>\n",
       "      <td>1.269444e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.015034e+03</td>\n",
       "      <td>5.248592e+04</td>\n",
       "      <td>6.865558e+00</td>\n",
       "      <td>5.248592e+04</td>\n",
       "      <td>4.098026e+03</td>\n",
       "      <td>5.436911e+02</td>\n",
       "      <td>48.115714</td>\n",
       "      <td>49.835901</td>\n",
       "      <td>70.660012</td>\n",
       "      <td>89.405494</td>\n",
       "      <td>72.782793</td>\n",
       "      <td>89.012859</td>\n",
       "      <td>107.807774</td>\n",
       "      <td>6.348572e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>7.940521e-01</td>\n",
       "      <td>2.641263e+04</td>\n",
       "      <td>3.857180e+00</td>\n",
       "      <td>2.641263e+04</td>\n",
       "      <td>6.122271e+04</td>\n",
       "      <td>2.945916e+02</td>\n",
       "      <td>30.422891</td>\n",
       "      <td>29.691985</td>\n",
       "      <td>44.620130</td>\n",
       "      <td>57.699279</td>\n",
       "      <td>48.232537</td>\n",
       "      <td>59.139094</td>\n",
       "      <td>73.241174</td>\n",
       "      <td>9.011435e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2.014000e+03</td>\n",
       "      <td>1.004600e+04</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.004600e+04</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.173000e+01</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.400000e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.014000e+03</td>\n",
       "      <td>3.021900e+04</td>\n",
       "      <td>4.000000e+00</td>\n",
       "      <td>3.021900e+04</td>\n",
       "      <td>2.933000e+01</td>\n",
       "      <td>3.391200e+02</td>\n",
       "      <td>26.710000</td>\n",
       "      <td>27.390000</td>\n",
       "      <td>27.910000</td>\n",
       "      <td>29.410000</td>\n",
       "      <td>28.260000</td>\n",
       "      <td>28.260000</td>\n",
       "      <td>29.490000</td>\n",
       "      <td>8.730000e+02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.015000e+03</td>\n",
       "      <td>4.953200e+04</td>\n",
       "      <td>6.000000e+00</td>\n",
       "      <td>4.953200e+04</td>\n",
       "      <td>2.916000e+02</td>\n",
       "      <td>4.632900e+02</td>\n",
       "      <td>49.770000</td>\n",
       "      <td>51.240000</td>\n",
       "      <td>74.680000</td>\n",
       "      <td>98.800000</td>\n",
       "      <td>77.250000</td>\n",
       "      <td>97.910000</td>\n",
       "      <td>116.710000</td>\n",
       "      <td>2.728000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.016000e+03</td>\n",
       "      <td>7.652600e+04</td>\n",
       "      <td>9.000000e+00</td>\n",
       "      <td>7.652600e+04</td>\n",
       "      <td>4.789800e+02</td>\n",
       "      <td>6.843900e+02</td>\n",
       "      <td>69.900000</td>\n",
       "      <td>72.230000</td>\n",
       "      <td>107.670000</td>\n",
       "      <td>137.080000</td>\n",
       "      <td>110.430000</td>\n",
       "      <td>136.330000</td>\n",
       "      <td>167.960000</td>\n",
       "      <td>7.577000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.016000e+03</td>\n",
       "      <td>9.996900e+04</td>\n",
       "      <td>2.400000e+01</td>\n",
       "      <td>9.996900e+04</td>\n",
       "      <td>9.999990e+05</td>\n",
       "      <td>6.604610e+03</td>\n",
       "      <td>182.400000</td>\n",
       "      <td>169.650000</td>\n",
       "      <td>251.740000</td>\n",
       "      <td>357.940000</td>\n",
       "      <td>260.850000</td>\n",
       "      <td>339.310000</td>\n",
       "      <td>449.140000</td>\n",
       "      <td>6.349300e+04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       BusinessYear      IssuerId    VersionNum     IssuerId2  IndividualRate  \\\n",
       "count  1.269444e+07  1.269444e+07  1.269444e+07  1.269444e+07    1.269444e+07   \n",
       "mean   2.015034e+03  5.248592e+04  6.865558e+00  5.248592e+04    4.098026e+03   \n",
       "std    7.940521e-01  2.641263e+04  3.857180e+00  2.641263e+04    6.122271e+04   \n",
       "min    2.014000e+03  1.004600e+04  1.000000e+00  1.004600e+04    0.000000e+00   \n",
       "25%    2.014000e+03  3.021900e+04  4.000000e+00  3.021900e+04    2.933000e+01   \n",
       "50%    2.015000e+03  4.953200e+04  6.000000e+00  4.953200e+04    2.916000e+02   \n",
       "75%    2.016000e+03  7.652600e+04  9.000000e+00  7.652600e+04    4.789800e+02   \n",
       "max    2.016000e+03  9.996900e+04  2.400000e+01  9.996900e+04    9.999990e+05   \n",
       "\n",
       "       IndividualTobaccoRate        Couple  PrimarySubscriberAndOneDependent  \\\n",
       "count           4.932349e+06  40941.000000                      40941.000000   \n",
       "mean            5.436911e+02     48.115714                         49.835901   \n",
       "std             2.945916e+02     30.422891                         29.691985   \n",
       "min             4.173000e+01      0.000000                          0.000000   \n",
       "25%             3.391200e+02     26.710000                         27.390000   \n",
       "50%             4.632900e+02     49.770000                         51.240000   \n",
       "75%             6.843900e+02     69.900000                         72.230000   \n",
       "max             6.604610e+03    182.400000                        169.650000   \n",
       "\n",
       "       PrimarySubscriberAndTwoDependents  \\\n",
       "count                       40941.000000   \n",
       "mean                           70.660012   \n",
       "std                            44.620130   \n",
       "min                             0.000000   \n",
       "25%                            27.910000   \n",
       "50%                            74.680000   \n",
       "75%                           107.670000   \n",
       "max                           251.740000   \n",
       "\n",
       "       PrimarySubscriberAndThreeOrMoreDependents  CoupleAndOneDependent  \\\n",
       "count                               40941.000000           40941.000000   \n",
       "mean                                   89.405494              72.782793   \n",
       "std                                    57.699279              48.232537   \n",
       "min                                     0.000000               0.000000   \n",
       "25%                                    29.410000              28.260000   \n",
       "50%                                    98.800000              77.250000   \n",
       "75%                                   137.080000             110.430000   \n",
       "max                                   357.940000             260.850000   \n",
       "\n",
       "       CoupleAndTwoDependents  CoupleAndThreeOrMoreDependents     RowNumber  \n",
       "count            40941.000000                    40941.000000  1.269444e+07  \n",
       "mean                89.012859                      107.807774  6.348572e+03  \n",
       "std                 59.139094                       73.241174  9.011435e+03  \n",
       "min                  0.000000                        0.000000  1.400000e+01  \n",
       "25%                 28.260000                       29.490000  8.730000e+02  \n",
       "50%                 97.910000                      116.710000  2.728000e+03  \n",
       "75%                136.330000                      167.960000  7.577000e+03  \n",
       "max                339.310000                      449.140000  6.349300e+04  "
      ]
     },
     "execution_count": 299,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9967748885437685"
      ]
     },
     "execution_count": 300,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Couple Rate NaN's out of whole data set (over 12 million rows!)\n",
    "12653504/12694445"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given that group rate columns have around 99.6% missing data, will only compare different individual rates across time and states since there is sufficient data to work with in those columns."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cleaning Individual Tobacco Rate data column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to handle NaN's, the choices are to replace NaN's with 0's or to take out entire rows that have NaN's. Our first approach will be the former."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate.IndividualTobaccoRate = Rate.IndividualTobaccoRate.fillna(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check to see if there are any more NaN's."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 302,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate.IndividualTobaccoRate.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are none. Great!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that the data is clean, let us take a look at individual costs with and without tobacco use summary statistics across time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualRate</th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualTobaccoRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>3796388.0</td>\n",
       "      <td>12922.257001</td>\n",
       "      <td>111452.391035</td>\n",
       "      <td>0.0</td>\n",
       "      <td>32.11</td>\n",
       "      <td>296.43</td>\n",
       "      <td>480.14</td>\n",
       "      <td>999999.00</td>\n",
       "      <td>3796388.0</td>\n",
       "      <td>240.926268</td>\n",
       "      <td>352.947647</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>427.64</td>\n",
       "      <td>6604.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4676092.0</td>\n",
       "      <td>329.161861</td>\n",
       "      <td>525.002167</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.43</td>\n",
       "      <td>283.38</td>\n",
       "      <td>469.69</td>\n",
       "      <td>9999.99</td>\n",
       "      <td>4676092.0</td>\n",
       "      <td>198.409085</td>\n",
       "      <td>304.982986</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>367.77</td>\n",
       "      <td>2379.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>4221965.0</td>\n",
       "      <td>337.539257</td>\n",
       "      <td>536.053504</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.49</td>\n",
       "      <td>296.30</td>\n",
       "      <td>488.07</td>\n",
       "      <td>9999.00</td>\n",
       "      <td>4221965.0</td>\n",
       "      <td>198.780763</td>\n",
       "      <td>310.452054</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>376.96</td>\n",
       "      <td>2677.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualRate                                                   \\\n",
       "                      count          mean            std  min    25%     50%   \n",
       "BusinessYear                                                                   \n",
       "2014              3796388.0  12922.257001  111452.391035  0.0  32.11  296.43   \n",
       "2015              4676092.0    329.161861     525.002167  0.0  28.43  283.38   \n",
       "2016              4221965.0    337.539257     536.053504  0.0  28.49  296.30   \n",
       "\n",
       "                                IndividualTobaccoRate                          \\\n",
       "                 75%        max                 count        mean         std   \n",
       "BusinessYear                                                                    \n",
       "2014          480.14  999999.00             3796388.0  240.926268  352.947647   \n",
       "2015          469.69    9999.99             4676092.0  198.409085  304.982986   \n",
       "2016          488.07    9999.00             4221965.0  198.780763  310.452054   \n",
       "\n",
       "                                              \n",
       "              min  25%  50%     75%      max  \n",
       "BusinessYear                                  \n",
       "2014          0.0  0.0  0.0  427.64  6604.61  \n",
       "2015          0.0  0.0  0.0  367.77  2379.86  \n",
       "2016          0.0  0.0  0.0  376.96  2677.79  "
      ]
     },
     "execution_count": 303,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate[[\"BusinessYear\",\"IndividualRate\",\"IndividualTobaccoRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There seems to be outliers in IndividualRate column since there is one value coming out at 999999.00. Also for IndividualToboccooRate 25% and 75% quartiles, 0's are what fill that portion of the dataset. Let us start dealing with these issues by eliminating outliers. Let us target our elimination areas by first looking at each individual rate's boxplots to see the distribution of data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1e714d42e8>"
      ]
     },
     "execution_count": 304,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualRate\", data=Rate)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This boxplot rendering is skewed because the data is that skewed. Outlier at 999999.00 is cause for the box to look like a line. All the data points are concentrated near the bottom because the gap between most of the points and the outlier is huge. The box plot will look different after outliers are removed so that data points will be closer to the Median. Another thing to note is that most outliers are in 2014 year for this column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1e714c0cc0>"
      ]
     },
     "execution_count": 305,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualTobaccoRate\", data=Rate)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For Ind. Rate for tobacco users, year 2014 has a huge amount of outliers that need to be eliminated. It looks like 2014 has the biggest data integrity issues."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will use zcore to drop outliers and perform it multiple times depending on how spread out data is. \n",
    "\n",
    "Zscore: In this procedure, z-score is calculated for each observation. Any z-score greater than +3 is considered to be an outlier. This rule of thumb is based on the empirical rule. From this rule we see that almost all of the data (99.7%) should be within three standard deviations from the mean. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate2 = Rate[(np.abs(stats.zscore(Rate.drop([\n",
    " 'BusinessYear',\n",
    " 'StateCode',\n",
    " 'IssuerId',\n",
    " 'SourceName',\n",
    " 'VersionNum',\n",
    " 'ImportDate',\n",
    " 'IssuerId2',\n",
    " 'FederalTIN',\n",
    " 'RateEffectiveDate',\n",
    " 'RateExpirationDate',\n",
    " 'PlanId',\n",
    " 'RatingAreaId',\n",
    " 'Tobacco',\n",
    " 'Age',\n",
    " 'Couple',\n",
    " 'PrimarySubscriberAndOneDependent',\n",
    " 'PrimarySubscriberAndTwoDependents',\n",
    " 'PrimarySubscriberAndThreeOrMoreDependents',\n",
    " 'CoupleAndOneDependent',\n",
    " 'CoupleAndTwoDependents',\n",
    " 'CoupleAndThreeOrMoreDependents',\n",
    " 'RowNumber'], axis=1))) < 3).all(axis=1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1bf0d5cdd8>"
      ]
     },
     "execution_count": 307,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualRate\", data=Rate2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>3674245.0</td>\n",
       "      <td>322.347511</td>\n",
       "      <td>487.618002</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31.44</td>\n",
       "      <td>286.89</td>\n",
       "      <td>453.32</td>\n",
       "      <td>9999.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4625636.0</td>\n",
       "      <td>321.025899</td>\n",
       "      <td>521.744693</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.25</td>\n",
       "      <td>279.72</td>\n",
       "      <td>460.69</td>\n",
       "      <td>9999.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>4174144.0</td>\n",
       "      <td>329.057729</td>\n",
       "      <td>532.930493</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.27</td>\n",
       "      <td>292.22</td>\n",
       "      <td>478.23</td>\n",
       "      <td>9999.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualRate                                              \\\n",
       "                      count        mean         std  min    25%     50%   \n",
       "BusinessYear                                                              \n",
       "2014              3674245.0  322.347511  487.618002  0.0  31.44  286.89   \n",
       "2015              4625636.0  321.025899  521.744693  0.0  28.25  279.72   \n",
       "2016              4174144.0  329.057729  532.930493  0.0  28.27  292.22   \n",
       "\n",
       "                               \n",
       "                 75%      max  \n",
       "BusinessYear                   \n",
       "2014          453.32  9999.99  \n",
       "2015          460.69  9999.99  \n",
       "2016          478.23  9999.00  "
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate2[[\"BusinessYear\",\"IndividualRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even after one elimination of values outside of +3 SD, there are still outliers. There are values at 9999.99. Another outlier elimination using z-score will be done to eliminate these values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate3 = Rate2[(np.abs(stats.zscore(Rate2.drop([\n",
    " 'BusinessYear',\n",
    " 'StateCode',\n",
    " 'IssuerId',\n",
    " 'SourceName',\n",
    " 'VersionNum',\n",
    " 'ImportDate',\n",
    " 'IssuerId2',\n",
    " 'FederalTIN',\n",
    " 'RateEffectiveDate',\n",
    " 'RateExpirationDate',\n",
    " 'PlanId',\n",
    " 'RatingAreaId',\n",
    " 'Tobacco',\n",
    " 'Age',\n",
    " 'Couple',\n",
    " 'PrimarySubscriberAndOneDependent',\n",
    " 'PrimarySubscriberAndTwoDependents',\n",
    " 'PrimarySubscriberAndThreeOrMoreDependents',\n",
    " 'CoupleAndOneDependent',\n",
    " 'CoupleAndTwoDependents',\n",
    " 'CoupleAndThreeOrMoreDependents',\n",
    " 'RowNumber'], axis=1))) < 3).all(axis=1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b7c13aef0>"
      ]
     },
     "execution_count": 310,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualRate\", data=Rate3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>3626007.0</td>\n",
       "      <td>298.956321</td>\n",
       "      <td>263.935910</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31.13</td>\n",
       "      <td>283.09</td>\n",
       "      <td>443.69</td>\n",
       "      <td>1867.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4571893.0</td>\n",
       "      <td>295.416474</td>\n",
       "      <td>281.852432</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.01</td>\n",
       "      <td>275.79</td>\n",
       "      <td>451.56</td>\n",
       "      <td>1871.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>4122018.0</td>\n",
       "      <td>302.087733</td>\n",
       "      <td>288.024715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.00</td>\n",
       "      <td>287.88</td>\n",
       "      <td>468.10</td>\n",
       "      <td>1870.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualRate                                              \\\n",
       "                      count        mean         std  min    25%     50%   \n",
       "BusinessYear                                                              \n",
       "2014              3626007.0  298.956321  263.935910  0.0  31.13  283.09   \n",
       "2015              4571893.0  295.416474  281.852432  0.0  28.01  275.79   \n",
       "2016              4122018.0  302.087733  288.024715  0.0  28.00  287.88   \n",
       "\n",
       "                               \n",
       "                 75%      max  \n",
       "BusinessYear                   \n",
       "2014          443.69  1867.91  \n",
       "2015          451.56  1871.07  \n",
       "2016          468.10  1870.62  "
      ]
     },
     "execution_count": 311,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate3[[\"BusinessYear\",\"IndividualRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yes! The extreme values have been eliminated from IndividualRate column. They were most likely due to input error. Now moving onto IndividualTocabbocRate column. Let us check with boxplot to see how we can remedy 0 values at 25% and 75% quartiles. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b35a93c18>"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualTobaccoRate\", data=Rate3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks like it is. Now, let us verify with the summary statistics table that min, median, max, 25%, 50% and 75% values are not as spread out as it was in original dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualRate</th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualTobaccoRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>3626007.0</td>\n",
       "      <td>298.956321</td>\n",
       "      <td>263.935910</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31.13</td>\n",
       "      <td>283.09</td>\n",
       "      <td>443.69</td>\n",
       "      <td>1867.91</td>\n",
       "      <td>3626007.0</td>\n",
       "      <td>208.191792</td>\n",
       "      <td>281.530171</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>403.10</td>\n",
       "      <td>1058.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4571893.0</td>\n",
       "      <td>295.416474</td>\n",
       "      <td>281.852432</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.01</td>\n",
       "      <td>275.79</td>\n",
       "      <td>451.56</td>\n",
       "      <td>1871.07</td>\n",
       "      <td>4571893.0</td>\n",
       "      <td>177.253162</td>\n",
       "      <td>267.452153</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>349.71</td>\n",
       "      <td>1058.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>4122018.0</td>\n",
       "      <td>302.087733</td>\n",
       "      <td>288.024715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.00</td>\n",
       "      <td>287.88</td>\n",
       "      <td>468.10</td>\n",
       "      <td>1870.62</td>\n",
       "      <td>4122018.0</td>\n",
       "      <td>176.357773</td>\n",
       "      <td>271.709137</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>357.74</td>\n",
       "      <td>1058.44</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualRate                                              \\\n",
       "                      count        mean         std  min    25%     50%   \n",
       "BusinessYear                                                              \n",
       "2014              3626007.0  298.956321  263.935910  0.0  31.13  283.09   \n",
       "2015              4571893.0  295.416474  281.852432  0.0  28.01  275.79   \n",
       "2016              4122018.0  302.087733  288.024715  0.0  28.00  287.88   \n",
       "\n",
       "                              IndividualTobaccoRate                          \\\n",
       "                 75%      max                 count        mean         std   \n",
       "BusinessYear                                                                  \n",
       "2014          443.69  1867.91             3626007.0  208.191792  281.530171   \n",
       "2015          451.56  1871.07             4571893.0  177.253162  267.452153   \n",
       "2016          468.10  1870.62             4122018.0  176.357773  271.709137   \n",
       "\n",
       "                                              \n",
       "              min  25%  50%     75%      max  \n",
       "BusinessYear                                  \n",
       "2014          0.0  0.0  0.0  403.10  1058.44  \n",
       "2015          0.0  0.0  0.0  349.71  1058.44  \n",
       "2016          0.0  0.0  0.0  357.74  1058.44  "
      ]
     },
     "execution_count": 313,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate3[[\"BusinessYear\",\"IndividualRate\",\"IndividualTobaccoRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "IndividualRate summary statistics looks good. However, IndividualTobaccoRate 25% and 75% values remain at 0. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This must mean where 25% and 50% of the data fall, it must mostly be 0. Let us verify this with bar chart/line graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b2ca96080>"
      ]
     },
     "execution_count": 314,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Rate3.IndividualTobaccoRate.hist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This confirms it. Most of the values are 0, but not necessarily $0. This was due to all the NaN's (7762096/12694445 = .61) that were replaced with 0's. In order to just handle the actual rate data that was given, will create a smaller dataframe that includes original data minus NaN's. This will be done prior to visualization of IndividualTobaccoRate across across states."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## c. Individual Rates Across Time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>3626007.0</td>\n",
       "      <td>298.956321</td>\n",
       "      <td>263.935910</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31.13</td>\n",
       "      <td>283.09</td>\n",
       "      <td>443.69</td>\n",
       "      <td>1867.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4571893.0</td>\n",
       "      <td>295.416474</td>\n",
       "      <td>281.852432</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.01</td>\n",
       "      <td>275.79</td>\n",
       "      <td>451.56</td>\n",
       "      <td>1871.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>4122018.0</td>\n",
       "      <td>302.087733</td>\n",
       "      <td>288.024715</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.00</td>\n",
       "      <td>287.88</td>\n",
       "      <td>468.10</td>\n",
       "      <td>1870.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualRate                                              \\\n",
       "                      count        mean         std  min    25%     50%   \n",
       "BusinessYear                                                              \n",
       "2014              3626007.0  298.956321  263.935910  0.0  31.13  283.09   \n",
       "2015              4571893.0  295.416474  281.852432  0.0  28.01  275.79   \n",
       "2016              4122018.0  302.087733  288.024715  0.0  28.00  287.88   \n",
       "\n",
       "                               \n",
       "                 75%      max  \n",
       "BusinessYear                   \n",
       "2014          443.69  1867.91  \n",
       "2015          451.56  1871.07  \n",
       "2016          468.10  1870.62  "
      ]
     },
     "execution_count": 315,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate3[[\"BusinessYear\",\"IndividualRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see here that most Individual Rates were stable across three years. However, to see the breakdown of each Individual Rate, let us do a count of each price that was offered in 10 different ranges and sorted in descending order. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.8719999999999999, 187.107]    4934499\n",
       "(187.107, 374.214]                3060551\n",
       "(374.214, 561.321]                2270829\n",
       "(561.321, 748.428]                1113097\n",
       "(748.428, 935.535]                 527613\n",
       "(935.535, 1122.642]                352648\n",
       "(1122.642, 1309.749]                42136\n",
       "(1309.749, 1496.856]                13223\n",
       "(1496.856, 1683.963]                 4209\n",
       "(1683.963, 1871.07]                  1113\n",
       "Name: IndividualRate, dtype: int64"
      ]
     },
     "execution_count": 316,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate3['IndividualRate'].value_counts(bins=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Therefore, for half of the dataset, most Individual Rate costs for non-tobacco users are 0 - 374 dollars from 2014 to 2016."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let us graph all IndividualRates per state to see if different rates are correlated with different states. Want to also see if there is variation in rates within each state. So this will be a two-dimensional assessment (across x-axis and across y-axis)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## d. Individual Rates Across States and Within each State using Boxplots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b25715208>"
      ]
     },
     "execution_count": 317,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15, 6))\n",
    "sns.set_style(\"whitegrid\")\n",
    "sns.boxplot(x=\"StateCode\", y=\"IndividualRate\", data=Rate3, order=Statelist)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above graph, we see that Arkansas has the most expensive rates whereas Montana has the cheapest based on where the median layed in each box plot. Regarding variability, Hawaii has the least variable whereas Texas has the most variable set of prices for Individual Rates. These differneces may be due to a state having mostly older people so variation is not high. With states that have greater variation with prices, this is most likely due to a state being big like Texas where age population is would be more diverse."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let us move onto Individual Tobacco Rates across time and states."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## a. More Data Integrity/Cleaning of Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As mentioned before, will only work with data that has no NaN's (eliminated rather than changed to 0 value). In order to get the necessary data points, will make a smaller dataframe from original dataset, with relevant columns to make computations easier (only include Business Year, StateCode and IndividualTobaccoRate). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Need to reload dataset and make a copy of it.\n",
    "Rate = pd.read_csv(\"Rate.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate4 = Rate[['BusinessYear', 'StateCode', 'IndividualTobaccoRate']].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 12694445 entries, 0 to 12694444\n",
      "Data columns (total 3 columns):\n",
      "BusinessYear             int64\n",
      "StateCode                object\n",
      "IndividualTobaccoRate    float64\n",
      "dtypes: float64(1), int64(1), object(1)\n",
      "memory usage: 290.6+ MB\n"
     ]
    }
   ],
   "source": [
    "Rate4.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Strategy for cleaning up IndividualTobaccoRate column: Reminder of how many NaN's, Remove NaN's, remove outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7762096"
      ]
     },
     "execution_count": 321,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate4['IndividualTobaccoRate'].isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NaNs: 61.15%\n"
     ]
    }
   ],
   "source": [
    "a = str(round(((7762096/12694445)*100),2))\n",
    "\n",
    "print('NaNs: ' + a + '%')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is a huge amount of data being eliminated. So there will possibly be a lot of time and state data lost as this huge chunk of data is eliminated from the whole dataset. At the same time, do not want to incorporate $0 amount as a valid price."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (b. More Summary Statistics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    4.932349e+06\n",
       "mean     5.436911e+02\n",
       "std      2.945916e+02\n",
       "min      4.173000e+01\n",
       "25%      3.391200e+02\n",
       "50%      4.632900e+02\n",
       "75%      6.843900e+02\n",
       "max      6.604610e+03\n",
       "Name: IndividualTobaccoRate, dtype: float64"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate4[\"IndividualTobaccoRate\"].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Confirmation: No 0's in original column. So all 0's from before were coming from NaN's being changed into 0's. Now, we will eliminate all NaN's to get data manipulations and visualizations on all available Ind. Tob. Rate data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Eliminate all NaN's from Ind. Tob. Rate column.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate5 = Rate4.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BusinessYear             0\n",
       "StateCode                0\n",
       "IndividualTobaccoRate    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 325,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate5.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    4.932349e+06\n",
       "mean     5.436911e+02\n",
       "std      2.945916e+02\n",
       "min      4.173000e+01\n",
       "25%      3.391200e+02\n",
       "50%      4.632900e+02\n",
       "75%      6.843900e+02\n",
       "max      6.604610e+03\n",
       "Name: IndividualTobaccoRate, dtype: float64"
      ]
     },
     "execution_count": 326,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate5['IndividualTobaccoRate'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset went from 12.7 million to 4.9 million. There is still a significant amount of data left to work with even though took out 60% of it. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 327,
   "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>BusinessYear</th>\n",
       "      <th>StateCode</th>\n",
       "      <th>IndividualTobaccoRate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14551</th>\n",
       "      <td>2014</td>\n",
       "      <td>AL</td>\n",
       "      <td>111.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14552</th>\n",
       "      <td>2014</td>\n",
       "      <td>AL</td>\n",
       "      <td>193.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14555</th>\n",
       "      <td>2014</td>\n",
       "      <td>AL</td>\n",
       "      <td>193.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14556</th>\n",
       "      <td>2014</td>\n",
       "      <td>AL</td>\n",
       "      <td>193.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14559</th>\n",
       "      <td>2014</td>\n",
       "      <td>AL</td>\n",
       "      <td>193.59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       BusinessYear StateCode  IndividualTobaccoRate\n",
       "14551          2014        AL                 111.74\n",
       "14552          2014        AL                 193.59\n",
       "14555          2014        AL                 193.59\n",
       "14556          2014        AL                 193.59\n",
       "14559          2014        AL                 193.59"
      ]
     },
     "execution_count": 327,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate5.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let us clean the data by eliminating the outliers using z-score. Let us remind ourselves of the current data spread with boxplots and summary statistics."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22b2848048>"
      ]
     },
     "execution_count": 328,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualTobaccoRate\", data=Rate5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualTobaccoRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>1634426.0</td>\n",
       "      <td>559.615174</td>\n",
       "      <td>333.179319</td>\n",
       "      <td>55.66</td>\n",
       "      <td>341.30</td>\n",
       "      <td>475.65</td>\n",
       "      <td>694.80</td>\n",
       "      <td>6604.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>1765295.0</td>\n",
       "      <td>525.566058</td>\n",
       "      <td>272.844149</td>\n",
       "      <td>60.59</td>\n",
       "      <td>327.91</td>\n",
       "      <td>448.46</td>\n",
       "      <td>665.35</td>\n",
       "      <td>2379.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>1532628.0</td>\n",
       "      <td>547.585862</td>\n",
       "      <td>272.948212</td>\n",
       "      <td>41.73</td>\n",
       "      <td>349.65</td>\n",
       "      <td>467.72</td>\n",
       "      <td>693.99</td>\n",
       "      <td>2677.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualTobaccoRate                                         \\\n",
       "                             count        mean         std    min     25%   \n",
       "BusinessYear                                                                \n",
       "2014                     1634426.0  559.615174  333.179319  55.66  341.30   \n",
       "2015                     1765295.0  525.566058  272.844149  60.59  327.91   \n",
       "2016                     1532628.0  547.585862  272.948212  41.73  349.65   \n",
       "\n",
       "                                       \n",
       "                 50%     75%      max  \n",
       "BusinessYear                           \n",
       "2014          475.65  694.80  6604.61  \n",
       "2015          448.46  665.35  2379.86  \n",
       "2016          467.72  693.99  2677.79  "
      ]
     },
     "execution_count": 329,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate5[[\"BusinessYear\",\"IndividualTobaccoRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [],
   "source": [
    "Rate6 = Rate5[(np.abs(stats.zscore(Rate5.drop([\n",
    " 'BusinessYear',\n",
    " 'StateCode'], axis=1))) < 3).all(axis=1)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us boxplot in order to see Ind. Tob. Rate column's distribution around Median."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a583cfc50>"
      ]
     },
     "execution_count": 331,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(x=\"BusinessYear\", y=\"IndividualTobaccoRate\", data=Rate6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Boxes from box plots have clear quartiles without extreme values as it looked like in year 2014 especially."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## c. Individual Tobacco Rates Across Time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"8\" halign=\"left\">IndividualTobaccoRate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BusinessYear</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>1604806.0</td>\n",
       "      <td>535.234569</td>\n",
       "      <td>262.911830</td>\n",
       "      <td>55.66</td>\n",
       "      <td>339.07</td>\n",
       "      <td>469.90</td>\n",
       "      <td>678.4575</td>\n",
       "      <td>1427.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>1753985.0</td>\n",
       "      <td>518.566027</td>\n",
       "      <td>259.013066</td>\n",
       "      <td>60.59</td>\n",
       "      <td>327.15</td>\n",
       "      <td>446.67</td>\n",
       "      <td>658.9900</td>\n",
       "      <td>1427.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>1523000.0</td>\n",
       "      <td>540.774562</td>\n",
       "      <td>259.511634</td>\n",
       "      <td>41.73</td>\n",
       "      <td>348.95</td>\n",
       "      <td>465.98</td>\n",
       "      <td>687.6200</td>\n",
       "      <td>1427.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             IndividualTobaccoRate                                         \\\n",
       "                             count        mean         std    min     25%   \n",
       "BusinessYear                                                                \n",
       "2014                     1604806.0  535.234569  262.911830  55.66  339.07   \n",
       "2015                     1753985.0  518.566027  259.013066  60.59  327.15   \n",
       "2016                     1523000.0  540.774562  259.511634  41.73  348.95   \n",
       "\n",
       "                                         \n",
       "                 50%       75%      max  \n",
       "BusinessYear                             \n",
       "2014          469.90  678.4575  1427.41  \n",
       "2015          446.67  658.9900  1427.45  \n",
       "2016          465.98  687.6200  1427.27  "
      ]
     },
     "execution_count": 332,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Rate6[[\"BusinessYear\",\"IndividualTobaccoRate\"]].groupby('BusinessYear').describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On average, the most expensive Ind. Tob. Rates came from 2016 with 2014 not too far behind. In comparison with Ind. Rates without Tobacco use, these rates are not as stable over time and are more expensive."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## d. Individual Tobacco Rate Across States and Within each State using Boxplots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23ad5775c0>"
      ]
     },
     "execution_count": 333,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15, 6))\n",
    "sns.set_style(\"whitegrid\")\n",
    "sns.boxplot(x=\"StateCode\", y=\"IndividualTobaccoRate\", data=Rate6, order=Statelist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 334,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "38 states are listed, which is one off number of states used to analyze Ind. Rates.\n"
     ]
    }
   ],
   "source": [
    "print(str(len(Rate6.StateCode.unique())) + \" states are listed, which is one off number of states used to analyze Ind. Rates.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we confirmed our data amount, will now comment on variability across states and variability within states."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are different Ind Tobacco users' rates. The highest are from AK, IN and WY. The cheapest are in UT, NM, KS and AL. The most variable Ind Tobacco Rates witin a state are KS and OR. These rates compared to non-tobacco users are more expensive by approximately $240."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Overall Lessons"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data integrity is key prior to working with such massive datasets as above. The huge amount of NaN's at times cautioned me from providing too general of summary statistics where the number of datapoints remaining did not provide enough solid representation of real life population data. At the same time, the dataset started at around 12 million and after the NaN's were taken out with Individual Tobacco Rate data, the amount of data remained massive. So general statements made from data still held weight. Also, huge gaps in data meant that multiple eliminations of data outside of z-score of +3 had to be done. This helped clear data of any anomalies that would affect general statements placed on data after analysis."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final Summary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Definitely, adoption of health plans vary across time and states. Number of plan carriers increased from 2014 to 2015 and then dropped slightly in 2016. Individual Rate plans for non-tobacco users were stable across time. On the other hand, Individual Rate plans for tobacco users decreased from 2014 to 2015 and increased again in 2016. And comparing pricing points between non-tobacco users and tobacco users, Individual Rates for tobacco users were more expensive as expected. Those that have higher health risks are paying for higher premiums. \n",
    "\n",
    "Given the wealth of data and having it span across multiple states, there is much more to be discovered especially in regards to gaps in state data and group rate data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Next Steps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If there was more resources and time, I would definitely evolve my data analysis by going into: \n",
    "    1. What sort of health insurance do people buy if they are heading into maturnity leave?\n",
    "    2. What sort of health insurance is bought when payers are under Medicare (>65 years old)?\n",
    "    3. Are there more purchases in a certain socioeconomic cut of America?\n",
    "    5. When are most purchases besides during open enrollment in November?\n",
    "    6. Who usually buys HMO/PPO?\n",
    "    7. How much of the US buys individual plans versus group plans?\n",
    "    8. Are there plans that are exclusive to states?"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}