Flatiron+Health+take+home.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df1 = pd.ExcelFile('/Users/harishprabhala/Downloads/flatiron_qs_orders_admins_july_16.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "orders = df1.parse('Orders')\n",
    "admins = df1.parse('Admins')\n",
    "dems = df1.parse('Demographics')\n",
    "patients = df1.parse('Patients')\n",
    "practice = df1.parse('Practices')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>order_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id   order_date\n",
       "0     h9993d                   70  w187808812917e  13-Dec-2012\n",
       "1     h9993d                   70   r90118525421e   7-Aug-2015\n",
       "2     h9993d                   70   h84358125584w  30-Sep-2015\n",
       "3     h9993d                   70   o81913333209g   8-Jan-2014\n",
       "4     h9993d                   70   e76139113344l  11-Oct-2012"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orders.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id administered_date  \\\n",
       "0     h9993d                   70  w187808812917e       13-Dec-2012   \n",
       "1     h9993d                   70   r90118525421e        7-Aug-2015   \n",
       "2     h9993d                   70   h84358125584w       30-Sep-2015   \n",
       "3     h9993d                   70   o81913333209g        8-Jan-2014   \n",
       "4     h9993d                   70   e76139113344l       11-Oct-2012   \n",
       "\n",
       "       drug_name  \n",
       "0    bevacizumab  \n",
       "1      nivolumab  \n",
       "2    bevacizumab  \n",
       "3      nivolumab  \n",
       "4  erlotinib hcl  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## 1.\tWhen presented with a new dataset or database, what steps do you generally take to evaluate it prior to working with it?  ",
    "\n",
    "When I am presented with a new dataset or database, prior to conducting the analysis, I generally look for/evaluate the following:\n",
    "\n",
    "- a.\tSource of the data – to make sure the source is reliable, valid and accurate\n",
    "- b.\tSize of the dataset – to evaluate which analytical tool to use that is optimal/efficient according to the size of the data.\n",
    "- c.\tUnderstand the data – to make sure that I understand the tables and columns so that there is no misinterpretation of the data.\n",
    "- d.\tData dictionary/documentation - request for a data dictionary if necessary/applicable.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Based on the information provided above and the attached dataset, what 3 questions would you like to understand prior to conducting any analysis of the data?  ",
    "\n",
    "\n",
    "I would like to understand the following three questions before conducting my analysis:\n",
    "\n",
    "- Demographic characteristics of the cancer patients\n",
    "- Drug/Medication usage\n",
    "- Patient volume per practice."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Demographic Characteristics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>race</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>female</td>\n",
       "      <td>69.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>female</td>\n",
       "      <td>46.0</td>\n",
       "      <td>WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>female</td>\n",
       "      <td>84.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>female</td>\n",
       "      <td>87.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>t9815h</td>\n",
       "      <td>female</td>\n",
       "      <td>97.0</td>\n",
       "      <td>WHITE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  gender   age       race\n",
       "0     h9993d  female  69.0  NON_WHITE\n",
       "1     w9986a  female  46.0      WHITE\n",
       "2     n9925d  female  84.0  NON_WHITE\n",
       "3     a9913g  female  87.0  NON_WHITE\n",
       "4     t9815h  female  97.0      WHITE"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dems.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "female     50\n",
       "male       43\n",
       "unknown     6\n",
       "Name: gender, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dems['gender'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x11636a908>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFJCAYAAAC2OXUDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsOzPd+YG1CcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11634d7b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=dems.gender.value_counts().index, y=dems.gender.value_counts())\n",
    "plt.ylabel('count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WHITE        57\n",
       "NON_WHITE    42\n",
       "Name: race, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dems['race'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1164a1748>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFJCAYAAAC2OXUDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQCH+L7qRbvmeOf1uAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116006e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=dems.race.value_counts().index, y=dems.race.value_counts())\n",
    "plt.ylabel('count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     99.000000\n",
       "mean      68.212121\n",
       "std       24.315223\n",
       "min       10.000000\n",
       "25%       50.500000\n",
       "50%       73.000000\n",
       "75%       86.000000\n",
       "max      118.000000\n",
       "Name: age, dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dems['age'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x115fe7160>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFXCAYAAACP5RboAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116026e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dems['age'].plot(kind='hist', bins=20, edgecolor='black')\n",
    "plt.xlabel('age')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Drug Usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id administered_date  \\\n",
       "0     h9993d                   70  w187808812917e       13-Dec-2012   \n",
       "1     h9993d                   70   r90118525421e        7-Aug-2015   \n",
       "2     h9993d                   70   h84358125584w       30-Sep-2015   \n",
       "3     h9993d                   70   o81913333209g        8-Jan-2014   \n",
       "4     h9993d                   70   e76139113344l       11-Oct-2012   \n",
       "\n",
       "       drug_name  \n",
       "0    bevacizumab  \n",
       "1      nivolumab  \n",
       "2    bevacizumab  \n",
       "3      nivolumab  \n",
       "4  erlotinib hcl  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "nivolumab        109\n",
       "bevacizumab      104\n",
       "erlotinib hcl     63\n",
       "axitinib          28\n",
       "Name: drug_name, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.drug_name.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1165c7a20>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfIAAAFJCAYAAABgnK9mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGqRJREFUeJzt3X10zGf+//HXJBFWJKiGVq0SZZXe0GrC7ko3lia0uj22\nqKzgOFXbY5exDlIimRK38aVlV9HWYamWLbtHf/Rm4y5bVhrWTVG3q5a4aUKsJJpkmOv3h2O+9dXK\nJGQmF8/HXzGT+XzeM1cmz/mMmYnDGGMEAACsFBToAQAAQOURcgAALEbIAQCwGCEHAMBihBwAAIsR\ncgAALBYS6AEqIy+vMNAjAADgN5GR4T94HkfkAABYjJADAGAxQg4AgMUIOQAAFiPkAABYjJADAGAx\nQg4AgMUIOQAAFiPkAABYjJADAGAxQg4AgMUIOQAAFiPkAABYzMq/flYRIzLWBHqEu8Kbo58P9AgA\ncFfiiBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbI\nAQCwGCEHAMBid/wfTYHdRv+/lECPcMfLeC490CMAuAUckQMAYDFCDgCAxQg5AAAWI+QAAFiMkAMA\nYDFCDgCAxao05Lt371ZSUpIk6fjx4+rXr58SExOVlpYmj8cjSVq5cqV69eqlPn36aOPGjVU5DgAA\nd5wqC/nbb7+tlJQUlZaWSpKmTp0qp9Op5cuXyxij9evXKy8vT0uXLtUHH3ygd999V7NmzVJZWVlV\njQQAwB2nykLetGlTzZ071/vvffv2KTo6WpIUGxurrVu3as+ePWrfvr1CQ0MVHh6upk2b6sCBA1U1\nEgAAd5wqC3l8fLxCQv73g+OMMXI4HJKksLAwFRYWqqioSOHh4d7vCQsLU1FRUVWNBADAHcdvH9Ea\nFPS/jxmKi4sVERGhOnXqqLi4+LrTvxv2H1K/fm2FhARXyZyonMjI8tcN1RNrB9jNbyFv06aNsrOz\nFRMTo6ysLHXs2FGPPfaY3njjDZWWlqqsrExHjx5Vq1atyt1WQcElP0yMisjLKwz0CKgk1g6o/m72\ngNtvIR87dqwmTJigWbNmKSoqSvHx8QoODlZSUpISExNljNHIkSNVs2ZNf40EAID1qjTkTZo00cqV\nKyVJzZs317Jly274nj59+qhPnz5VOQYAAHcsPhAGAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIA\nACxGyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQc\nAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBgh\nBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQcAACLEXIAACxG\nyAEAsBghBwDAYoQcAACLEXIAACxGyAEAsBghBwDAYoQcAACLhfhzZ263W8nJycrNzVVQUJAmTZqk\nkJAQJScny+FwqGXLlkpLS1NQEI8vAADwhV9DvnnzZl2+fFkffPCBtmzZojfeeENut1tOp1MxMTFK\nTU3V+vXr1a1bN3+OBQCAtfx66Nu8eXNduXJFHo9HRUVFCgkJ0b59+xQdHS1Jio2N1datW/05EgAA\nVvPrEXnt2rWVm5ur7t27q6CgQPPnz1dOTo4cDockKSwsTIWFheVup3792goJCa7qcVEBkZHhgR4B\nlcTaAXbza8gXL16sn//85xo1apROnz6tgQMHyu12e88vLi5WREREudspKLhUlWOiEvLyyn8AhuqJ\ntQOqv5s94PbrU+sREREKD786TN26dXX58mW1adNG2dnZkqSsrCx16NDBnyMBAGA1vx6RDxo0SOPG\njVNiYqLcbrdGjhypRx55RBMmTNCsWbMUFRWl+Ph4f44EAIDV/BrysLAwvfnmmzecvmzZMn+OAQDA\nHYM3bAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5\nAAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFC\nDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiM\nkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWCwn0AADuTDmjhgd6hLvCU/8zJ9Aj\nIMA4IgcAwGKEHAAAixFyAAAs5lPIJ02adMNpY8eOve3DAACAirnpi93Gjx+vEydOaO/evTp8+LD3\n9MuXL6uwsLBSO1ywYIE2bNggt9utfv36KTo6WsnJyXI4HGrZsqXS0tIUFMQTBQAA+OKmIX/11VeV\nm5uryZMn63e/+5339ODgYLVo0aLCO8vOztbOnTv1/vvv69tvv9WiRYs0depUOZ1OxcTEKDU1VevX\nr1e3bt0qfk0AALgL3TTkTZo0UZMmTbRmzRoVFRWpsLBQxhhJ0qVLl1SvXr0K7ezzzz9Xq1atNGzY\nMBUVFWnMmDFauXKloqOjJUmxsbHasmULIQcAwEc+vY98wYIFWrBgwXXhdjgcWr9+fYV2VlBQoFOn\nTmn+/Pk6efKkXn31VRlj5HA4JElhYWE+PWVfv35thYQEV2jfqFqRkeGBHgGVxNrZjfWDTyH/y1/+\noszMTN1zzz23tLN69eopKipKoaGhioqKUs2aNXXmzBnv+cXFxYqIiCh3OwUFl25pDtx+eXmVe80E\nAo+1sxvrd3e42QM2n15Vdv/996tu3bq3PMiTTz6pf/zjHzLG6OzZs/r222/VqVMnZWdnS5KysrLU\noUOHW94PAAB3C5+OyJs1a6bExETFxMQoNDTUe/p3XwDni7i4OOXk5OjFF1+UMUapqalq0qSJJkyY\noFmzZikqKkrx8fEVuwYAANzFfAp5o0aN1KhRo9uywzFjxtxw2rJly27LtgEAuNv4FPKKHnkDAAD/\n8CnkrVu39r6y/JqGDRtq8+bNVTIUAADwjU8hP3DggPdrt9utzMxM7dq1q8qGAgAAvqnwZ6HWqFFD\n3bt317Zt26piHgAAUAE+HZH/7W9/835tjNHhw4dVo0aNKhsKAAD4xqeQX3uf9zX169fX7Nmzq2Qg\nAADgO59CPnXqVLndbh07dkxXrlxRy5YtFRLi00UBAEAV8qnGe/fu1fDhw1WvXj15PB7l5+frT3/6\nkx5//PGqng8AANyETyFPT0/X7NmzveHetWuXJk2apA8//LBKhwMAADfn06vWL126dN3Rd7t27VRa\nWlplQwEAAN/4FPK6desqMzPT++/MzMwK/y1yAABw+/n01PqkSZM0dOhQjR8/3nvaBx98UGVDAQAA\n3/h0RJ6VlaUf/ehH2rhxo5YsWaJ77rlHX3zxRVXPBgAAyuFTyFeuXKn3339ftWvXVuvWrbV69Wr+\nYhkAANWATyF3u93XfZIbn+oGAED14NP/kXft2lUDBw5U9+7dJUmfffaZfvnLX1bpYAAAoHw+hXz0\n6NH65JNPlJOTo5CQEA0YMEBdu3at6tkAAEA5fP6c1YSEBCUkJFTlLAAAoIIq/GdMAQBA9UHIAQCw\nGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAA\nLEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwA\nAIsRcgAALEbIAQCwWEBCfu7cOT399NM6evSojh8/rn79+ikxMVFpaWnyeDyBGAkAACv5PeRut1up\nqamqVauWJGnq1KlyOp1avny5jDFav369v0cCAMBafg/59OnT9dJLL6lhw4aSpH379ik6OlqSFBsb\nq61bt/p7JAAArBXiz52tXr1a99xzjzp37qyFCxdKkowxcjgckqSwsDAVFhaWu5369WsrJCS4SmdF\nxURGhgd6BFQSa2c31g9+DfmqVavkcDj0z3/+U1999ZXGjh2r8+fPe88vLi5WREREudspKLhUlWOi\nEvLyyn8AhuqJtbMb63d3uNkDNr+G/L333vN+nZSUJJfLpYyMDGVnZysmJkZZWVnq2LGjP0cCAMBq\nAX/72dixYzV37lz17dtXbrdb8fHxgR4JAABr+PWI/LuWLl3q/XrZsmWBGgMAAKsF/IgcAABUHiEH\nAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbI\nAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsRcgAALEbIAQCwGCEHAMBihBwAAIsR\ncgAALEbIAQCwGCEHAMBihBwAAIuFBHoAAED18/YbnwR6hDveEGfCbdkOR+QAAFiMkAMAYDFCDgCA\nxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMAYDFCDgCAxQg5AAAWI+QAAFiMkAMA\nYDFCDgCAxQg5AAAWC/Hnztxut8aNG6fc3FyVlZXp1Vdf1UMPPaTk5GQ5HA61bNlSaWlpCgri8QUA\nAL7wa8jXrFmjevXqKSMjQxcuXNALL7yg1q1by+l0KiYmRqmpqVq/fr26devmz7EAALCWXw99ExIS\nNGLECEmSMUbBwcHat2+foqOjJUmxsbHaunWrP0cCAMBqfg15WFiY6tSpo6KiIg0fPlxOp1PGGDkc\nDu/5hYWF/hwJAACr+fWpdUk6ffq0hg0bpsTERPXs2VMZGRne84qLixUREVHuNurXr62QkOCqHBMV\nFBkZHugRUEmsnd1YP3vdrrXza8jz8/M1ePBgpaamqlOnTpKkNm3aKDs7WzExMcrKylLHjh3L3U5B\nwaWqHhUVlJfHMym2Yu3sxvrZqyJrd7Po+/Wp9fnz5+vixYuaN2+ekpKSlJSUJKfTqblz56pv375y\nu92Kj4/350gAAFjNr0fkKSkpSklJueH0ZcuW+XMMAADuGLxhGwAAixFyAAAsRsgBALAYIQcAwGKE\nHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAY\nIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAs\nRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAA\nixFyAAAsRsgBALAYIQcAwGKEHAAAixFyAAAsRsgBALAYIQcAwGKEHAAAi4UEegBJ8ng8crlcOnjw\noEJDQ5Wenq4HH3ww0GMBAFDtVYsj8szMTJWVlWnFihUaNWqUpk2bFuiRAACwQrUI+Y4dO9S5c2dJ\nUrt27bR3794ATwQAgB0cxhgT6CHGjx+vZ555Rk8//bQk6Re/+IUyMzMVElItnvkHAKDaqhZH5HXq\n1FFxcbH33x6Ph4gDAOCDahHyJ554QllZWZKkXbt2qVWrVgGeCAAAO1SLp9avvWr90KFDMsZoypQp\natGiRaDHAgCg2qsWIQcAAJVTLZ5aBwAAlUPIAQCwGCG/zSZPnqxTp05V6DJdunRRaWnpbZ/lZz/7\n2W3fpm1Wr16tmTNnVvl+KrPulVVVPy93grlz5+r999//wfNPnTqlDRs2SCp/zUaOHKmysjIlJyd7\nX4z7fSryM8ba3R7X1u7ChQv66KOPJEkLFy7Unj17yr1MeT8jNuI9XrfZ+PHjAz0CAoB1t8O2bdv0\n73//W126dCl3zWbPnu2nqVBR19YuOztbGzZsUM+ePfXKK6/4dJk7ESG/BatXr9bmzZtVUlKi//zn\nPxoyZIj++te/yuVyafTo0ZozZ46aNGmiTz75RNu3b9fw4cM1evRoFRUV6cqVKxoxYoQ6derk3V5y\ncrJ69Oih2NhYZWVlad26dZo2bZq6deum9u3b6+uvv1anTp1UWFioPXv2qHnz5srIyNChQ4c0bdo0\nXblyRQUFBXK5XHriiSdUVlamkSNH6vTp0/rJT34il8slh8MRwFssMHbt2qWBAweqqKhIv//971W7\ndm3Nnj1bwcHB+vGPf6yJEydq5MiRGjBggKKjo/Xll19q3rx5ysjI0Pjx41VYWKhvvvlGiYmJSkxM\n1O7duzVlyhR5PB41atRIM2fO1JAhQ+RyubR27Vrl5ORIkg4dOqSkpCQZY3TvvfeqX79+Onr0qFwu\nl5YuXaqePXuqQ4cOOnjwoKKiotSgQQNt375doaGhWrhwoc6dOyeXy6XS0lLl5eXJ6XSqa9eukqTU\n1FTl5uaqQYMGmj59umrVqhXIm9gv3G630tLSdPz4cXk8HjmdTsXExOi5555Ts2bNVKNGDUVFRXm/\nf9q0adqxY4ck6bnnnlP//v21cOFClZSUqH379lq8eLFcLpfWrVunkydP6ty5czp16pRee+01de7c\nWV26dNHHH38sSVq+fLneffddXblyRZMnT77hb0Hs3r1bgwcP1vnz59WvXz/17dtXGzdu1B//+EcZ\nY9S2bVu9/vrr/ruxLFVUVHTdfa5v375au3athg0bpocfflgDBw7UO++8ozFjxsjlcmn+/Pk6cOCA\nVqxYoZ07d6pHjx7Kz8+/4fdyr169lJSUJJfLJenqx4J//PHHKikpUUpKih577LHAXvFbZVBpq1at\nMoMHDzbGGHPs2DETHx9v+vfvb44cOWLee+89M3fuXGOMMUOGDDEHDx4006ZNM4sXLzbGGHPmzBkT\nFxdnPB6PiYuLMyUlJWbs2LFm8+bNxhhjNm/ebMaOHWuMMebhhx82ubm5pqyszLRr184cPnzYe7n/\n/ve/Zu3atebAgQPGGGPWrFljxo8fb4wx5pFHHjEnT540xhgzfPhwk5mZ6b8bp5pYtWqVefnll43H\n4zH5+fkmLi7OPPPMMyY/P98YY8zs2bPNihUrzKZNm0xycrIxxhiXy2U2bNhg9u7daz799FNjzNX1\n6tatmzHGmOeff94cOXLEGGPMypUrzd69e73rfs3GjRvN4MGDTWlpqZkzZ45Zvny5McaYI0eOmP79\n+xtjjImLizPbt283xhgTHx9vNm3aZIwx5je/+Y3Zv3+/2bJli9m2bZsxxpgdO3aYQYMGeS+3c+dO\nY4wx06dPN0uWLKmiW696ee+998yMGTOMMcacP3/e9OjRwxhz9fbYt2+fMcZ4b+sNGzaYYcOGGY/H\nY8rKysyLL75oDhw4YFatWmUyMjKMMca7ZnPmzDEpKSnGGGM+//xz7336u/fLBQsWGGOM2bRpkxk2\nbNh1c61atcoMGjTIeDwec+LECdO9e3fjdrtNXFyc9+ds4cKFJjc317tNfL/vu8+dOHHCPPvss2bQ\noEHe+8i1tdu2bZtxOp3GGOP9/fl9v5e/e5k5c+aYCRMmGGOMOXTokHnhhRf8fTVvO47Ib1Hr1q0l\nSffff7/Kysq8p/fs2VOJiYnq3bu3ioqK1KpVKx09elQ9e/aUJDVq1Eh16tTRuXPnvne75jvvCqxX\nr54aN24sSapdu7YeeughSVJ4eLhKS0vVsGFDzZs3T7Vq1VJxcbHq1KkjSWrcuLEeeOABSVL79u11\n7Nix23zt7fDkk0/K4XCoQYMGqlWrlk6ePCmn0ylJKikp0U9/+lO9+OKLysjI0IULF7R9+3alpKQo\nPz9fS5Ys0WeffaY6dero8uXLkqT8/Hzv5xz07t37hv3l5ORo/vz5eueddxQaGnrT2dq2bStJioiI\n8G4zIiJCpaWlioyM1FtvvaUPP/xQDofDu/8aNWqoXbt2kq5+mNKWLVtuw61U/R06dEg7duzw/j/o\n5cuXdf78eUlS8+bNr/veo0ePqkOHDnI4HKpRo4Yef/xxHT169Ae3/fDDD0uS7rvvvuvux9d06NBB\n0tX70YwZM244v02bNnI4HIqMjFRJSYkKCgoUERGhBg0aSJKGDBlSiWt897n33ntvuM81adJETzzx\nhHbt2qXY2FiftvNDv5eveeqppyRJLVu2VF5e3u27AgHCi91u0Q89VR0eHq5HHnlEU6dOVa9evSRJ\nLVq00Pbt2yVJZ8+e1cWLF1WvXj3vZUJDQ70/VPv37y93H9dMnjxZw4cP1/Tp09WqVSvvg4AzZ87o\nm2++kST961//UsuWLSt5Le325ZdfSpLy8vJUWlqqBx54QPPmzdPSpUv129/+Vh07dlRQUJASEhLk\ncrnUtWtXBQcHa9GiRWrXrp1mzpyphIQE7+3asGFDff3115KuvsDm73//u3df+/fv15QpUzR37lzv\nA6qaNWt613Xfvn3XzXaztX3zzTf1q1/9ShkZGYqJifHu3+1266uvvpIkbd++/a5Z16ioKD377LNa\nunSp3n77bSUkJHjvP0FB1/8qa9GihfdpdbfbrZ07d+rBBx9UUFCQPB7PDdsu7z527cHDD93e//fy\nDRo00MWLF3XhwgVJUnp6+k1fiIWrvu8+t2vXLh0+fFhPPfWUFi1adN333+p6Hjx40HuQZDOOyKtQ\n79699fLLL2vKlCmSpKFDh2rcuHH69NNPVVJSookTJ173mfK9e/fWuHHj9NFHH6lZs2Y+7+f555/X\niBEjFBERofvuu08FBQWSrh7Jp6en6+zZs2rfvr33j9LcbUpKSjRgwABdunRJ6enpunLlil555RUZ\nYxQWFuY9wvr1r3+trl276tNPP5UkxcXFKT09XevWrVN4eLiCg4NVVlam119/XePGjVNQUJAiIyM1\naNAg/fnPf5YkjR49WsHBwfrDH/4gSXr00UfVr18/OZ1O5eTkeI/AfZGQkKAZM2Zo4cKF161rjRo1\ntHTpUh0/flyNGzfWqFGjbufNVW299NJLSklJUf/+/VVUVKTExMQbAn5NXFycvvjiC/Xt21dut1sJ\nCQlq27atHA6H3nrrrQqtg3T1/8AHDBggh8PhvT/fTFBQkNLS0jR06FAFBQWpTZs2evTRRyu0z7vR\n/73PORwOvfbaa5o3b54aN26s3r17Kzo62vv9TZs21aFDh7R48eIK7efkyZMaMGCAysrKNHHixNt8\nLfyPT3YDAMBiPLUOAIDFCDkAABYj5AAAWIyQAwBgMUIOAIDFCDkAABYj5AAAWIyQAwBgsf8PAIJ/\nAvrj8O4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1164ac470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=admins.drug_name.value_counts().index, y=admins.drug_name.value_counts())\n",
    "plt.ylabel('count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "external_patient_id  drug_name    \n",
       "1                    axitinib         1\n",
       "                     nivolumab        2\n",
       "2                    axitinib         1\n",
       "                     bevacizumab      2\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        6\n",
       "3                    bevacizumab      1\n",
       "                     erlotinib hcl    2\n",
       "4                    bevacizumab      1\n",
       "                     erlotinib hcl    3\n",
       "                     nivolumab        2\n",
       "5                    axitinib         1\n",
       "                     nivolumab        2\n",
       "6                    axitinib         1\n",
       "                     erlotinib hcl    2\n",
       "7                    bevacizumab      1\n",
       "                     nivolumab        1\n",
       "8                    nivolumab        1\n",
       "9                    axitinib         1\n",
       "                     bevacizumab      2\n",
       "                     erlotinib hcl    2\n",
       "10                   nivolumab        2\n",
       "11                   nivolumab        3\n",
       "12                   axitinib         1\n",
       "13                   bevacizumab      1\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        2\n",
       "14                   axitinib         2\n",
       "                     nivolumab        2\n",
       "15                   bevacizumab      2\n",
       "                                     ..\n",
       "90                   bevacizumab      1\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        1\n",
       "91                   bevacizumab      1\n",
       "                     nivolumab        3\n",
       "92                   axitinib         1\n",
       "                     bevacizumab      1\n",
       "                     nivolumab        1\n",
       "93                   bevacizumab      1\n",
       "94                   erlotinib hcl    1\n",
       "                     nivolumab        3\n",
       "95                   bevacizumab      3\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        4\n",
       "96                   axitinib         1\n",
       "                     bevacizumab      1\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        2\n",
       "97                   bevacizumab      1\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        1\n",
       "98                   bevacizumab      1\n",
       "                     erlotinib hcl    1\n",
       "                     nivolumab        1\n",
       "99                   axitinib         1\n",
       "                     bevacizumab      1\n",
       "100                  bevacizumab      1\n",
       "                     nivolumab        1\n",
       "104                  bevacizumab      2\n",
       "                     erlotinib hcl    1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d = admins[(admins.drug_name == 'nivolumab')|(admins.drug_name == 'bevacizumab')|(admins.drug_name == 'erlotinib hcl')|(admins.drug_name == 'axitinib')]\n",
    "g = d.groupby([d.external_patient_id,'drug_name']).size()\n",
    "g"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x118dd2b70>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdkAAAGJCAYAAADGyUn1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQyhZAAAMoWQBADDk/wEp3f3CDXPWbAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118ab7e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g.unstack(0).plot(kind ='bar', legend=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### In the above graph, we can see the frequency/count of the drug dosage in the patients. Nivolumab and Bevacizumab have been administered more frequently than erlotinib hcl and axitinib"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Patient volume per practice"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>external_practice_id</th>\n",
       "      <th>internal_practice_id</th>\n",
       "      <th>diagnosis_date</th>\n",
       "      <th>advanced_diagnosis_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>46</td>\n",
       "      <td>ZE4G</td>\n",
       "      <td>4</td>\n",
       "      <td>15-Feb-2012</td>\n",
       "      <td>3-Jul-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>15</td>\n",
       "      <td>FUM35</td>\n",
       "      <td>5</td>\n",
       "      <td>5-Jul-1962</td>\n",
       "      <td>18-May-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7</td>\n",
       "      <td>16-Oct-1930</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7</td>\n",
       "      <td>16-Oct-1930</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id external_practice_id  internal_practice_id  \\\n",
       "0     h9993d                   70                  QZ1                     9   \n",
       "1     w9986a                   46                 ZE4G                     4   \n",
       "2     n9925d                   15                FUM35                     5   \n",
       "3     a9913g                    6               DSVN67                     7   \n",
       "4     a9913g                    6               DSVN67                     7   \n",
       "\n",
       "  diagnosis_date advanced_diagnosis_date  \n",
       "0    29-Mar-1992                     NaN  \n",
       "1    15-Feb-2012              3-Jul-2012  \n",
       "2     5-Jul-1962             18-May-2009  \n",
       "3    16-Oct-1930             14-Sep-1970  \n",
       "4    16-Oct-1930             14-Sep-1970  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "patients.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DSVN67    15\n",
       "ZE4G      15\n",
       "FZAB      14\n",
       "FEI30     12\n",
       "FI3040    12\n",
       "KNO54     11\n",
       "FUM35      8\n",
       "QZ1        7\n",
       "FZ3        4\n",
       "WZ4        3\n",
       "Name: external_practice_id, dtype: int64"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "patients.external_practice_id.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x11b3c3390>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFJCAYAAAC2OXUDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAFji/wDZ41BR3wXX7QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11b1dcb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x=patients.external_practice_id.value_counts().index, y=patients.external_practice_id.value_counts())\n",
    "plt.ylabel('patient count')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The above graph shows the number of patients at each practice/client."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "prac_drugs=patients.merge(admins,how='outer',left_on='patient_id',right_on='patient_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 315 entries, 0 to 314\n",
      "Data columns (total 10 columns):\n",
      "patient_id                 315 non-null object\n",
      "external_patient_id_x      312 non-null float64\n",
      "external_practice_id       312 non-null object\n",
      "internal_practice_id       312 non-null float64\n",
      "diagnosis_date             305 non-null object\n",
      "advanced_diagnosis_date    305 non-null object\n",
      "external_patient_id_y      315 non-null int64\n",
      "order_id                   315 non-null object\n",
      "administered_date          315 non-null object\n",
      "drug_name                  315 non-null object\n",
      "dtypes: float64(2), int64(1), object(7)\n",
      "memory usage: 27.1+ KB\n"
     ]
    }
   ],
   "source": [
    "prac_drugs.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id_x</th>\n",
       "      <th>external_practice_id</th>\n",
       "      <th>internal_practice_id</th>\n",
       "      <th>diagnosis_date</th>\n",
       "      <th>advanced_diagnosis_date</th>\n",
       "      <th>external_patient_id_y</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id_x external_practice_id  \\\n",
       "0     h9993d                   70.0                  QZ1   \n",
       "1     h9993d                   70.0                  QZ1   \n",
       "2     h9993d                   70.0                  QZ1   \n",
       "3     h9993d                   70.0                  QZ1   \n",
       "4     h9993d                   70.0                  QZ1   \n",
       "\n",
       "   internal_practice_id diagnosis_date advanced_diagnosis_date  \\\n",
       "0                   9.0    29-Mar-1992                     NaN   \n",
       "1                   9.0    29-Mar-1992                     NaN   \n",
       "2                   9.0    29-Mar-1992                     NaN   \n",
       "3                   9.0    29-Mar-1992                     NaN   \n",
       "4                   9.0    29-Mar-1992                     NaN   \n",
       "\n",
       "   external_patient_id_y        order_id administered_date      drug_name  \n",
       "0                     70  w187808812917e       13-Dec-2012    bevacizumab  \n",
       "1                     70   r90118525421e        7-Aug-2015      nivolumab  \n",
       "2                     70   h84358125584w       30-Sep-2015    bevacizumab  \n",
       "3                     70   o81913333209g        8-Jan-2014      nivolumab  \n",
       "4                     70   e76139113344l       11-Oct-2012  erlotinib hcl  "
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prac_drugs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "external_practice_id  drug_name    \n",
       "DSVN67                axitinib          6\n",
       "                      bevacizumab      11\n",
       "                      erlotinib hcl    13\n",
       "                      nivolumab        13\n",
       "FEI30                 axitinib          2\n",
       "                      bevacizumab      12\n",
       "                      erlotinib hcl     6\n",
       "                      nivolumab        11\n",
       "FI3040                axitinib          5\n",
       "                      bevacizumab      22\n",
       "                      erlotinib hcl     9\n",
       "                      nivolumab        10\n",
       "FUM35                 bevacizumab       3\n",
       "                      erlotinib hcl     6\n",
       "                      nivolumab         8\n",
       "FZ3                   axitinib          3\n",
       "                      bevacizumab       3\n",
       "                      erlotinib hcl     3\n",
       "                      nivolumab         1\n",
       "FZAB                  axitinib          3\n",
       "                      bevacizumab      18\n",
       "                      erlotinib hcl     8\n",
       "                      nivolumab        21\n",
       "KNO54                 axitinib          1\n",
       "                      bevacizumab      12\n",
       "                      erlotinib hcl     8\n",
       "                      nivolumab        11\n",
       "QZ1                   axitinib          2\n",
       "                      bevacizumab       6\n",
       "                      erlotinib hcl     3\n",
       "                      nivolumab        11\n",
       "WZ4                   axitinib          1\n",
       "                      bevacizumab       3\n",
       "                      erlotinib hcl     1\n",
       "                      nivolumab         7\n",
       "ZE4G                  axitinib          6\n",
       "                      bevacizumab      15\n",
       "                      erlotinib hcl     8\n",
       "                      nivolumab        20\n",
       "dtype: int64"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prac_drugs.groupby(['external_practice_id','drug_name']).size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "p = prac_drugs.groupby(['external_practice_id','drug_name']).size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11be3dcf8>"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd8AAAF1CAYAAACph+DQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11b770b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p.unstack(1).plot(kind ='bar')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The above graph shows the frequency of each administered drug at each practice/client."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. How would you prep the dataset provided for analysis? Please list steps taken and provide code used to prep the tables for analysis.\n",
    "\n",
    "The dataset given for this exercise is small and fairly neat w/o any missing data. So, I do not find much need to prep the data in this case. However, if the dataset is big with large number of columns and rows, then I would take the following steps:\n",
    "\n",
    "- Evaluate the important columns that are useful for the analysis and keep them while removing the rest of the columns that are not useful.\n",
    "- If the size of the missing data is large, then thoroughly evaluate the columns and rows to conduct data imputation.\n",
    "- If there are any outliers that are heavily skewing the data, then evaluate whether or not it is important/prudent to remove the outliers.\n",
    "- Conduct central tendency tests and evaulate the variable spreads.\n",
    "- If the analysis requires predictive modeling/machine learning, then conduct feature engineering of the dependent and independent variables to build and train different models.\n",
    "\n",
    "Since, this analysis does not require to conduct any statistical testing or predictive modeling, the data prep steps are pretty limited in this case and I will be using the data as it is."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using the datasets you have prepped, please answer the following questions and provide all code/output used. Be sure to include and explain any assumptions made when answering each question."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. What is the average time elapsed between a patient’s initial diagnosis date and a patient’s first treatment? Does this time vary by gender?\n",
    "\n",
    "- The assumption here is that the patient's first treatment date is the date when the first drug was administered. For this analysis, I have taken the minimun of each patient's administered date. Thus, it gives us the earliest date when the patient has taken a drug."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id administered_date  \\\n",
       "0     h9993d                   70  w187808812917e       13-Dec-2012   \n",
       "1     h9993d                   70   r90118525421e        7-Aug-2015   \n",
       "2     h9993d                   70   h84358125584w       30-Sep-2015   \n",
       "3     h9993d                   70   o81913333209g        8-Jan-2014   \n",
       "4     h9993d                   70   e76139113344l       11-Oct-2012   \n",
       "\n",
       "       drug_name  \n",
       "0    bevacizumab  \n",
       "1      nivolumab  \n",
       "2    bevacizumab  \n",
       "3      nivolumab  \n",
       "4  erlotinib hcl  "
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "admins_copy = admins[['patient_id','administered_date']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/harishprabhala/anaconda/lib/python3.6/site-packages/ipykernel/__main__.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  if __name__ == '__main__':\n"
     ]
    }
   ],
   "source": [
    "admins_copy['administered_date'] = pd.DatetimeIndex(admins_copy['administered_date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>administered_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>2012-12-13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>2015-08-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>2015-09-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>2014-01-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>2012-10-11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id administered_date\n",
       "0     h9993d        2012-12-13\n",
       "1     h9993d        2015-08-07\n",
       "2     h9993d        2015-09-30\n",
       "3     h9993d        2014-01-08\n",
       "4     h9993d        2012-10-11"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins_copy.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.DataFrame(admins_copy.groupby(['patient_id'],sort=False)['administered_date'].min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>administered_date</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>patient_id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>h9993d</th>\n",
       "      <td>2012-10-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>w9986a</th>\n",
       "      <td>2011-05-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n9925d</th>\n",
       "      <td>2013-07-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a9913g</th>\n",
       "      <td>2013-11-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>t9815h</th>\n",
       "      <td>2014-06-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d9750n</th>\n",
       "      <td>2013-01-24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>t9509b</th>\n",
       "      <td>2012-10-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>s9417f</th>\n",
       "      <td>2013-05-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b9305v</th>\n",
       "      <td>2014-01-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>p9302d</th>\n",
       "      <td>2014-01-28</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           administered_date\n",
       "patient_id                  \n",
       "h9993d            2012-10-11\n",
       "w9986a            2011-05-25\n",
       "n9925d            2013-07-11\n",
       "a9913g            2013-11-12\n",
       "t9815h            2014-06-04\n",
       "d9750n            2013-01-24\n",
       "t9509b            2012-10-02\n",
       "s9417f            2013-05-02\n",
       "b9305v            2014-01-14\n",
       "p9302d            2014-01-28"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [],
   "source": [
    "diag_treat_diff=patients.merge(data,how='outer',left_on='patient_id',right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>external_practice_id</th>\n",
       "      <th>internal_practice_id</th>\n",
       "      <th>diagnosis_date</th>\n",
       "      <th>advanced_diagnosis_date</th>\n",
       "      <th>administered_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>29-Mar-1992</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2012-10-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>46.0</td>\n",
       "      <td>ZE4G</td>\n",
       "      <td>4.0</td>\n",
       "      <td>15-Feb-2012</td>\n",
       "      <td>3-Jul-2012</td>\n",
       "      <td>2011-05-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>15.0</td>\n",
       "      <td>FUM35</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5-Jul-1962</td>\n",
       "      <td>18-May-2009</td>\n",
       "      <td>2013-07-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16-Oct-1930</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16-Oct-1930</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id external_practice_id  internal_practice_id  \\\n",
       "0     h9993d                 70.0                  QZ1                   9.0   \n",
       "1     w9986a                 46.0                 ZE4G                   4.0   \n",
       "2     n9925d                 15.0                FUM35                   5.0   \n",
       "3     a9913g                  6.0               DSVN67                   7.0   \n",
       "4     a9913g                  6.0               DSVN67                   7.0   \n",
       "\n",
       "  diagnosis_date advanced_diagnosis_date administered_date  \n",
       "0    29-Mar-1992                     NaN        2012-10-11  \n",
       "1    15-Feb-2012              3-Jul-2012        2011-05-25  \n",
       "2     5-Jul-1962             18-May-2009        2013-07-11  \n",
       "3    16-Oct-1930             14-Sep-1970        2013-11-12  \n",
       "4    16-Oct-1930             14-Sep-1970        2013-11-12  "
      ]
     },
     "execution_count": 246,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diag_treat_diff.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "diag_treat_diff['diagnosis_date'] = pd.DatetimeIndex(diag_treat_diff['diagnosis_date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [],
   "source": [
    "diag_treat_diff['datediff'] = (diag_treat_diff['administered_date'] - diag_treat_diff['diagnosis_date']).dt.days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>external_practice_id</th>\n",
       "      <th>internal_practice_id</th>\n",
       "      <th>diagnosis_date</th>\n",
       "      <th>advanced_diagnosis_date</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>datediff</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1992-03-29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2012-10-11</td>\n",
       "      <td>7501.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>46.0</td>\n",
       "      <td>ZE4G</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2012-02-15</td>\n",
       "      <td>3-Jul-2012</td>\n",
       "      <td>2011-05-25</td>\n",
       "      <td>-266.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>15.0</td>\n",
       "      <td>FUM35</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1962-07-05</td>\n",
       "      <td>18-May-2009</td>\n",
       "      <td>2013-07-11</td>\n",
       "      <td>18634.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1930-10-16</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "      <td>30343.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1930-10-16</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "      <td>30343.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id external_practice_id  internal_practice_id  \\\n",
       "0     h9993d                 70.0                  QZ1                   9.0   \n",
       "1     w9986a                 46.0                 ZE4G                   4.0   \n",
       "2     n9925d                 15.0                FUM35                   5.0   \n",
       "3     a9913g                  6.0               DSVN67                   7.0   \n",
       "4     a9913g                  6.0               DSVN67                   7.0   \n",
       "\n",
       "  diagnosis_date advanced_diagnosis_date administered_date  datediff  \n",
       "0     1992-03-29                     NaN        2012-10-11    7501.0  \n",
       "1     2012-02-15              3-Jul-2012        2011-05-25    -266.0  \n",
       "2     1962-07-05             18-May-2009        2013-07-11   18634.0  \n",
       "3     1930-10-16             14-Sep-1970        2013-11-12   30343.0  \n",
       "4     1930-10-16             14-Sep-1970        2013-11-12   30343.0  "
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This table gives us the number of days elapsed (datediff) between the diagnosis date \n",
    "# and the first drug administered date \n",
    "diag_treat_diff.head() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### In the above table, there are some outliers where the administered date is earlier than the initial diagnosis date. These may be mistakenly entered so we omit these rows for our analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "external_patient_id        51.239583\n",
       "internal_practice_id        5.447917\n",
       "datediff                14622.281250\n",
       "dtype: float64"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diag_treat_diff[(diag_treat_diff.datediff>0)].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The average days elapsed between initial diagnosis and first treatment is 14622 days or 40 years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "diag_treat_diff_gender=diag_treat_diff.merge(dems,how='outer',left_on='patient_id',right_on='patient_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>external_practice_id</th>\n",
       "      <th>internal_practice_id</th>\n",
       "      <th>diagnosis_date</th>\n",
       "      <th>advanced_diagnosis_date</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>datediff</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>race</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70.0</td>\n",
       "      <td>QZ1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1992-03-29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2012-10-11</td>\n",
       "      <td>7501.0</td>\n",
       "      <td>female</td>\n",
       "      <td>69.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>46.0</td>\n",
       "      <td>ZE4G</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2012-02-15</td>\n",
       "      <td>3-Jul-2012</td>\n",
       "      <td>2011-05-25</td>\n",
       "      <td>-266.0</td>\n",
       "      <td>female</td>\n",
       "      <td>46.0</td>\n",
       "      <td>WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>15.0</td>\n",
       "      <td>FUM35</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1962-07-05</td>\n",
       "      <td>18-May-2009</td>\n",
       "      <td>2013-07-11</td>\n",
       "      <td>18634.0</td>\n",
       "      <td>female</td>\n",
       "      <td>84.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1930-10-16</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "      <td>30343.0</td>\n",
       "      <td>female</td>\n",
       "      <td>87.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>6.0</td>\n",
       "      <td>DSVN67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1930-10-16</td>\n",
       "      <td>14-Sep-1970</td>\n",
       "      <td>2013-11-12</td>\n",
       "      <td>30343.0</td>\n",
       "      <td>female</td>\n",
       "      <td>87.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id external_practice_id  internal_practice_id  \\\n",
       "0     h9993d                 70.0                  QZ1                   9.0   \n",
       "1     w9986a                 46.0                 ZE4G                   4.0   \n",
       "2     n9925d                 15.0                FUM35                   5.0   \n",
       "3     a9913g                  6.0               DSVN67                   7.0   \n",
       "4     a9913g                  6.0               DSVN67                   7.0   \n",
       "\n",
       "  diagnosis_date advanced_diagnosis_date administered_date  datediff  gender  \\\n",
       "0     1992-03-29                     NaN        2012-10-11    7501.0  female   \n",
       "1     2012-02-15              3-Jul-2012        2011-05-25    -266.0  female   \n",
       "2     1962-07-05             18-May-2009        2013-07-11   18634.0  female   \n",
       "3     1930-10-16             14-Sep-1970        2013-11-12   30343.0  female   \n",
       "4     1930-10-16             14-Sep-1970        2013-11-12   30343.0  female   \n",
       "\n",
       "    age       race  \n",
       "0  69.0  NON_WHITE  \n",
       "1  46.0      WHITE  \n",
       "2  84.0  NON_WHITE  \n",
       "3  87.0  NON_WHITE  \n",
       "4  87.0  NON_WHITE  "
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diag_treat_diff_gender.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "external_patient_id        53.250000\n",
       "internal_practice_id        5.875000\n",
       "datediff                16047.479167\n",
       "age                        70.958333\n",
       "dtype: float64"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diag_treat_diff_gender[(diag_treat_diff_gender.datediff>0)&(diag_treat_diff_gender.gender == 'female')].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### For females, the average days elapsed between initial diagnosis and first treatment is 16047 days or 43.9 years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "external_patient_id        48.333333\n",
       "internal_practice_id        5.000000\n",
       "datediff                13017.285714\n",
       "age                        62.023810\n",
       "dtype: float64"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diag_treat_diff_gender[(diag_treat_diff_gender.datediff>0)&(diag_treat_diff_gender.gender == 'male')].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### For males, the average days elapsed between initial diagnosis and first treatment is 13017 days or 35.6 years"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. How many patients are on nivolumab from 2012 - 2016?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>13-Dec-2012</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>7-Aug-2015</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>30-Sep-2015</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>8-Jan-2014</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>11-Oct-2012</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id administered_date  \\\n",
       "0     h9993d                   70  w187808812917e       13-Dec-2012   \n",
       "1     h9993d                   70   r90118525421e        7-Aug-2015   \n",
       "2     h9993d                   70   h84358125584w       30-Sep-2015   \n",
       "3     h9993d                   70   o81913333209g        8-Jan-2014   \n",
       "4     h9993d                   70   e76139113344l       11-Oct-2012   \n",
       "\n",
       "       drug_name  \n",
       "0    bevacizumab  \n",
       "1      nivolumab  \n",
       "2    bevacizumab  \n",
       "3      nivolumab  \n",
       "4  erlotinib hcl  "
      ]
     },
     "execution_count": 257,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "admins['administered_date'] = pd.DatetimeIndex(admins['administered_date'] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100"
      ]
     },
     "execution_count": 278,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(admins[(admins.administered_date >='2012-01-01')&(admins.administered_date <'2016-01-01')\n",
    "      & (admins.drug_name == 'nivolumab')])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Between 2012 and 2016, there were 100 patients on nivolumab"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Using the following risk stratification rules, please summarize the number of high, medium, and low risk patients.\n",
    "- HighRisk = Female; any­age; NON_WHITE OR Male; >= 70; NON_WHITE \n",
    "- MediumRisk = Female; >=75, WHITE OR Male; <70; NON_WHITE\n",
    "- LowRisk = Female; < 75, WHITE OR Male; any­age; WHITE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>gender</th>\n",
       "      <th>age</th>\n",
       "      <th>race</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>female</td>\n",
       "      <td>69.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>w9986a</td>\n",
       "      <td>female</td>\n",
       "      <td>46.0</td>\n",
       "      <td>WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>n9925d</td>\n",
       "      <td>female</td>\n",
       "      <td>84.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a9913g</td>\n",
       "      <td>female</td>\n",
       "      <td>87.0</td>\n",
       "      <td>NON_WHITE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>t9815h</td>\n",
       "      <td>female</td>\n",
       "      <td>97.0</td>\n",
       "      <td>WHITE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  gender   age       race\n",
       "0     h9993d  female  69.0  NON_WHITE\n",
       "1     w9986a  female  46.0      WHITE\n",
       "2     n9925d  female  84.0  NON_WHITE\n",
       "3     a9913g  female  87.0  NON_WHITE\n",
       "4     t9815h  female  97.0      WHITE"
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dems.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of high risk patients are 30\n"
     ]
    }
   ],
   "source": [
    "high_risk = (len(dems[(dems.gender=='female')&(dems.race=='NON_WHITE')]))+(len(dems[(dems.gender=='male')&(dems.race=='NON_WHITE')&(dems.age>=70)]))\n",
    "print('The number of high risk patients are %d' %high_risk) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of medium risk patients are 24\n"
     ]
    }
   ],
   "source": [
    "med_risk = (len(dems[(dems.gender=='female')&(dems.age>=75)&(dems.race=='WHITE')]))+(len(dems[(dems.gender=='male')&(dems.race=='NON_WHITE')&(dems.age<70)]))\n",
    "print('The number of medium risk patients are %d' %med_risk) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The number of low risk patients are 39\n"
     ]
    }
   ],
   "source": [
    "low_risk = (len(dems[(dems.gender=='female')&(dems.age<75)&(dems.race=='WHITE')]))+(len(dems[(dems.gender=='male')&(dems.race=='WHITE')]))\n",
    "print('The number of low risk patients are %d' %low_risk) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Please create a visualization that could be used to help a medical researcher? understand how drug prevalence has changed over time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patient_id</th>\n",
       "      <th>external_patient_id</th>\n",
       "      <th>order_id</th>\n",
       "      <th>administered_date</th>\n",
       "      <th>drug_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>w187808812917e</td>\n",
       "      <td>2012-12-13</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>r90118525421e</td>\n",
       "      <td>2015-08-07</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>h84358125584w</td>\n",
       "      <td>2015-09-30</td>\n",
       "      <td>bevacizumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>o81913333209g</td>\n",
       "      <td>2014-01-08</td>\n",
       "      <td>nivolumab</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>h9993d</td>\n",
       "      <td>70</td>\n",
       "      <td>e76139113344l</td>\n",
       "      <td>2012-10-11</td>\n",
       "      <td>erlotinib hcl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  patient_id  external_patient_id        order_id administered_date  \\\n",
       "0     h9993d                   70  w187808812917e        2012-12-13   \n",
       "1     h9993d                   70   r90118525421e        2015-08-07   \n",
       "2     h9993d                   70   h84358125584w        2015-09-30   \n",
       "3     h9993d                   70   o81913333209g        2014-01-08   \n",
       "4     h9993d                   70   e76139113344l        2012-10-11   \n",
       "\n",
       "       drug_name  \n",
       "0    bevacizumab  \n",
       "1      nivolumab  \n",
       "2    bevacizumab  \n",
       "3      nivolumab  \n",
       "4  erlotinib hcl  "
      ]
     },
     "execution_count": 299,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admins.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "admins_copy1 = admins[['patient_id','administered_date','drug_name']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/harishprabhala/anaconda/lib/python3.6/site-packages/ipykernel/__main__.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  if __name__ == '__main__':\n"
     ]
    }
   ],
   "source": [
    "admins_copy1['admin_year'] = admins_copy1['administered_date'].dt.year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [],
   "source": [
    "admins_copy1 = admins_copy1[admins_copy1.admin_year < 2018]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [],
   "source": [
    "drg = admins_copy1.groupby(['admin_year','drug_name']).size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11d3cb198>"
      ]
     },
     "execution_count": 319,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd8AAAFkCAYAAABhDTHuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11d1319b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAd8AAAFXCAYAAADj40TtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhBBeJuErhBBCeNn/B1RWBDqigR1zAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11d318a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "drg.unstack(1).plot(kind='Bar')\n",
    "drg.unstack(1).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The above two graphs represent the change of drug prevalence over time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}