COVID19_INDIA_PATIENT_CLUSTER_ANALYSIS.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import networkx as nx\n",
    "from collections import OrderedDict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### COVID 19 INDIA CROWD-SOURCED DATBASE \n",
    "Source - covid19india.org\n",
    "<br>\n",
    "Their data sources - https://telegra.ph/Covid-19-Sources-03-19"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"COVID19_raw_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.fillna(\"-1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(23041, 20)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = [data.columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_filtered = data.loc[data['Contracted from which Patient (Suspected)']!='-1']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1534, 20)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_filtered.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "def extractor(x):\n",
    "    if x[0]=='E' and x[1]=='0':\n",
    "        return [1670]\n",
    "    elif x[0]=='P':\n",
    "        numbers = re.findall('\\d+',x)\n",
    "        numbers = [int(i) for i in numbers]\n",
    "        if numbers!=None:\n",
    "            return numbers\n",
    "        else:\n",
    "            return []\n",
    "    else:\n",
    "        return []\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ashish-j/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.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",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "data_filtered['source_cluster_flag'] = data_filtered['Contracted from which Patient (Suspected)'].apply(lambda x:extractor(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "nodes = list(data_filtered['source_cluster_flag'].index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "adjacency_list = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from pyvis.network import Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_graph(adjacency_list):\n",
    "    g = nx.from_dict_of_lists(adjacency_list)\n",
    "    g1 = Network(height='750',width='750',notebook=True,bgcolor='#000000')\n",
    "    options = '''{\n",
    "      \"physics\": {\n",
    "        \"forceAtlas2Based\": {\n",
    "          \"gravitationalConstant\": -547,\n",
    "          \"springLength\": 275,\n",
    "          \"avoidOverlap\": 0.98\n",
    "        },\n",
    "        \"maxVelocity\": 82,\n",
    "        \"minVelocity\": 0.87,\n",
    "        \"solver\": \"forceAtlas2Based\"\n",
    "      }\n",
    "    }'''\n",
    "    g1.set_options(options)\n",
    "    g1.from_nx(g)\n",
    "    return g1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def assign_node_color(nodes,color_map,clr):\n",
    "    adjacency_list ={}\n",
    "    color_nodes = OrderedDict()\n",
    "    for i in nodes:\n",
    "        if clr==True:\n",
    "            state = data_filtered['Detected State'][i]\n",
    "            if state in color_map.keys():\n",
    "                adjacency_list[i+1]=data_filtered['source_cluster_flag'][i]\n",
    "                assign_color = color_map[state]\n",
    "                color_nodes[i+1] = assign_color\n",
    "        elif clr==False:\n",
    "            adjacency_list[i+1]=data_filtered['source_cluster_flag'][i]      \n",
    "    return color_nodes,adjacency_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "def graph_color(gl,color_nodes,label,states):\n",
    "    #print(color_nodes)\n",
    "    for i in range(len(gl.nodes)):\n",
    "        try:\n",
    "            gl.nodes[i]['color']=color_nodes[int(gl.nodes[i]['title'])]\n",
    "            gl.nodes[i]['size']=40\n",
    "            \n",
    "        except: \n",
    "            state = data.iloc[int(gl.nodes[i]['title'])-1]['Detected State']\n",
    "            if(len(states)!=0):\n",
    "                if state==states[0]:\n",
    "                    gl.nodes[i]['color']=\"#ff0000\"\n",
    "                    gl.nodes[i]['size']=40\n",
    "                    #print(\"red-ed\")\n",
    "                elif state==states[1]:\n",
    "                    gl.nodes[i]['color']=\"#0000ff\"\n",
    "                    gl.nodes[i]['size']=40\n",
    "                    #print(\"blued\")\n",
    "                else:\n",
    "                    gl.nodes[i]['color']=\"#000000\"\n",
    "                    gl.nodes[i]['size']=40\n",
    "                    \n",
    "            else:\n",
    "                gl.nodes[i]['color']=\"#3b3b3b\"\n",
    "                gl.nodes[i]['size']=40\n",
    "    gl.prep_notebook()\n",
    "    gl.show(\"./interactive_graphs/covid_19_\"+str(label)+\".html\")\n",
    "def just_show_graph(gl,label):\n",
    "    gl.show(\"./interactive_graphs/covid_19_\"+str(label)+\".html\")\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### ** Nodes represent 'Patient Number' in data, download covid_19.html for interactive graph. Here central cluster - travelled to Delhi cases.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Naive Analysis - 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/naive_analysis.jpeg\">\n",
    "We are looking such kind of patterns where recepient becomes source of infection for others as well.\n",
    "for e.g p3 got infected from p1 and then p5 and p6 got infected from p3.\n",
    "<b> Higher the number, longer and denser the chain of such patterns->higher is the chance that community transmission is happening<b> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map = {}\n",
    "_,adjacency_list_full = assign_node_color(nodes,color_map,False)\n",
    "gr = build_graph(adjacency_list_full)\n",
    "just_show_graph(gr,\"FULL\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### examples of case described above in data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/full.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) Neighbhouring State Spread Analysis\n",
    "<b>Two of more than 2 different colors in single conected component of graph represents some kind og inter-state spread most commonly observed in  \"Travelled to Delhi\"<b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1) Delhi - UP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map_up_delhi = {'Uttar Pradesh':\"#ff0000\",'Delhi':\"#0000ff\"}\n",
    "color_nodes_up_delhi,adjacency_list_up_delhi = assign_node_color(nodes,color_map_up_delhi,True)\n",
    "gr = build_graph(adjacency_list_up_delhi)\n",
    "graph_color(gr,color_nodes_up_delhi,\"UP_DELHI\",[\"Uttar Pradesh\",\"Delhi\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> red = UP , blue = Delhi <b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/up_delhi.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2) Maharshtra - Madhya Pradesh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map_mah_madh = {'Maharashtra':\"#ff0000\",'Madhya Pradesh':\"#0000ff\"}\n",
    "color_nodes_mah_madh,adjacency_list_mah_madh = assign_node_color(nodes,color_map_mah_madh,True)\n",
    "gr = build_graph(adjacency_list_mah_madh)\n",
    "graph_color(gr,color_nodes_mah_madh,\"MAHA_MP\",['Maharashtra','Madhya Pradesh'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> red = Maharashtra , blue = MP <b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/maha_mp.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3)Kerala - Tamil Nadu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map_kl_tam = {'Kerala':\"#ff0000\",'Tamil Nadu':\"#0000ff\"}\n",
    "color_nodes_kl_tam,adjacency_list_kl_tam = assign_node_color(nodes,color_map_kl_tam,True)\n",
    "gr = build_graph(adjacency_list_kl_tam)\n",
    "graph_color(gr,color_nodes_kl_tam,\"KL_TAMIL\",['Kerala','Tamil Nadu'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> red = Kerala , blue = Tamil Nadu <b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/kl_tamil.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4) Gujarat - Rajasthan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map_guj_raj = {'Gujarat':\"#ff0000\",'Rajasthan':\"#0000ff\"}\n",
    "color_nodes_guj_raj,adjacency_list_guj_raj = assign_node_color(nodes,color_map_guj_raj,True)\n",
    "gr = build_graph(adjacency_list_guj_raj)\n",
    "graph_color(gr,color_nodes_guj_raj,\"GUJ_RAJ\",['Gujarat','Rajasthan'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> red = Gujarat , blue = Rajasthan <b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/guj_raj.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 5) TOP 9 state-wise spread (Source node of Central cluster represents \"Travelled to Delhi\" not a person"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "color_map_misc = {\"Kerala\":\"#ff0000\",                #red\n",
    "            \"Tamil Nadu\" : \"#0000ff\",               #blue\n",
    "            \"Maharashtra\" : \"#09d712\",              #bright green\n",
    "            \"Delhi\" : \"#ccb080\",                    #light brown\n",
    "            \"Uttar Pradesh\":\"#ff83fa\",               #purple\n",
    "            \"Madhya Pradesh\" : \"#6a8000\",            #brwonish yellow\n",
    "            \"Gujarat\":\"#ffd700\",                    #yellow orangish\"\n",
    "            \"Rajasthan\":\"#2cdede\",                   #cyan\n",
    "            \"Karnataka\":\"#dadada\"}                   #grey\n",
    "\n",
    "color_nodes_misc,adjacency_list_misc = assign_node_color(nodes,color_map_misc,True)\n",
    "gr = build_graph(adjacency_list_misc)\n",
    "graph_color(gr,color_nodes_misc,\"TOP_8_STATES\",[])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/top_8.png\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}