distance_info.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>In this file we transform data on player location and parachute landing to get the distances between players at different moments in time along the game.\n",
    "NOT FINISHED</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import json\n",
    "import time\n",
    "import datetime\n",
    "from importlib import reload\n",
    "\n",
    "import requests\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.cluster import DBSCAN\n",
    "\n",
    "import external as ext"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Defining API calls info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define api key\n",
    "apikey = \"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJqdGkiOiI4MjFjMGY4MC00NmFlLTAxMzgtYTkwMi0yNWIyZjA5NmRmYWMiLCJpc3MiOiJnYW1lbG9ja2VyIiwiaWF0IjoxNTg0MDMxNDAxLCJwdWIiOiJibHVlaG9sZSIsInRpdGxlIjoicHViZyIsImFwcCI6Im1pZ3VlbC1oZXJyYWV6In0.FvFkXnOYxXMtXeArAATb3f_BFrA3NgE9GBuTT6iL3fs\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define request header\n",
    "headers = {\n",
    "    \"Authorization\": f'Bearer {apikey}',\n",
    "    \"Accept\": \"application/vnd.api+json\"\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the area\n",
    "area = \"pc-eu\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting match ids"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=5px>This part has been moved to the \"extract_data.ipynb\" notebook</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting match data (single match)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p><font size=5px>This part has been moved to the \"extract_data.ipynb\" notebook</font></p>\n",
    "<p><font size=4px>What is now here is for the purpose of the analysis of the data of this single match</font></p>\n",
    "<p>From here onwards, we are using the data on match 4c76d82b-90f7-4fd1-9ceb-97b99347e94e, belonging to the pc shard</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# collect the match data of one match. We need this to get the assets id, needed to get the full telemetry data\n",
    "url_collectassetsid = f\"https://api.pubg.com/shards/pc/matches/4c76d82b-90f7-4fd1-9ceb-97b99347e94e\"\n",
    "response2 = requests.get(url_collectassetsid, headers=headers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "#response2.json()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting match telemetry data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p><font size=5px>This part has been moved to the \"extract_data.ipynb\" notebook</font></p>\n",
    "<p><font size=4px>What is now here is for the purpose of the analysis of the data of this single match</font></p>\n",
    "<p>From here onwards, we are using the data on match 4c76d82b-90f7-4fd1-9ceb-97b99347e94e, belonging to the pc shard</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# UNDERSTANDING HOW EVENTS WORK\n",
    "# get kills for the match (will be used to understand how location works after death)\n",
    "kills = []\n",
    "for i in json.loads(json.dumps(telemetry_response.json())):\n",
    "    if i[\"_T\"] == \"LogPlayerKill\":\n",
    "        kills.append(i)\n",
    "# kills"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# <font color=red> Preparing DataFrame </font>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# RUN THIS CELL TO WORK WITH THE FILE UPLOADED IN THE REPOSITORY\n",
    "# read files extracted from this as DataFrame\n",
    "with open('telemetry_data/2020-04-27/console/0cc1afa0-cfb4-4de8-86c2-7075a34d92d5.json', 'r') as file:\n",
    "    json_data = file.read()\n",
    "    file.close()\n",
    "data = []\n",
    "for i in json.loads(json_data):\n",
    "    if i[\"_T\"] in (\"LogPlayerPosition\", \"LogParachuteLanding\"):\n",
    "        data.append(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>character</th>\n",
       "      <th>vehicle</th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>numAlivePlayers</th>\n",
       "      <th>common</th>\n",
       "      <th>_D</th>\n",
       "      <th>_T</th>\n",
       "      <th>distance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>{'name': 'heyyzeus_', 'teamId': 3, 'health': 1...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.701Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>{'name': 'BERRYWEBBLE3000', 'teamId': 10, 'hea...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.731Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>{'name': 'walas-ii', 'teamId': 1, 'health': 10...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:26.263Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>{'name': 'Zensity v', 'teamId': 11, 'health': ...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.195Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>{'name': 'Anerge', 'teamId': 11, 'health': 100...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.196Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           character vehicle  elapsedTime  \\\n",
       "0  {'name': 'heyyzeus_', 'teamId': 3, 'health': 1...    None          0.0   \n",
       "1  {'name': 'BERRYWEBBLE3000', 'teamId': 10, 'hea...    None          0.0   \n",
       "2  {'name': 'walas-ii', 'teamId': 1, 'health': 10...    None          0.0   \n",
       "3  {'name': 'Zensity v', 'teamId': 11, 'health': ...    None          0.0   \n",
       "4  {'name': 'Anerge', 'teamId': 11, 'health': 100...    None          0.0   \n",
       "\n",
       "   numAlivePlayers         common                        _D  \\\n",
       "0             51.0  {'isGame': 0}  2020-04-26T07:58:25.701Z   \n",
       "1             51.0  {'isGame': 0}  2020-04-26T07:58:25.731Z   \n",
       "2             52.0  {'isGame': 0}  2020-04-26T07:58:26.263Z   \n",
       "3             57.0  {'isGame': 0}  2020-04-26T07:58:27.195Z   \n",
       "4             57.0  {'isGame': 0}  2020-04-26T07:58:27.196Z   \n",
       "\n",
       "                  _T  distance  \n",
       "0  LogPlayerPosition       NaN  \n",
       "1  LogPlayerPosition       NaN  \n",
       "2  LogPlayerPosition       NaN  \n",
       "3  LogPlayerPosition       NaN  \n",
       "4  LogPlayerPosition       NaN  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get events data into pandas dataframe\n",
    "data = pd.read_json(json.dumps(data))\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 8526 entries, 0 to 8525\n",
      "Data columns (total 8 columns):\n",
      "character          8526 non-null object\n",
      "vehicle            1320 non-null object\n",
      "elapsedTime        8428 non-null float64\n",
      "numAlivePlayers    8428 non-null float64\n",
      "common             8526 non-null object\n",
      "_D                 8526 non-null object\n",
      "_T                 8526 non-null object\n",
      "distance           98 non-null float64\n",
      "dtypes: float64(3), object(5)\n",
      "memory usage: 533.0+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>character</th>\n",
       "      <th>vehicle</th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>numAlivePlayers</th>\n",
       "      <th>common</th>\n",
       "      <th>_D</th>\n",
       "      <th>_T</th>\n",
       "      <th>distance</th>\n",
       "      <th>MatchId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>{'name': 'heyyzeus_', 'teamId': 3, 'health': 1...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.701Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>{'name': 'BERRYWEBBLE3000', 'teamId': 10, 'hea...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.731Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>{'name': 'walas-ii', 'teamId': 1, 'health': 10...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:26.263Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>{'name': 'Zensity v', 'teamId': 11, 'health': ...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.195Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>{'name': 'Anerge', 'teamId': 11, 'health': 100...</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.196Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           character vehicle  elapsedTime  \\\n",
       "0  {'name': 'heyyzeus_', 'teamId': 3, 'health': 1...    None          0.0   \n",
       "1  {'name': 'BERRYWEBBLE3000', 'teamId': 10, 'hea...    None          0.0   \n",
       "2  {'name': 'walas-ii', 'teamId': 1, 'health': 10...    None          0.0   \n",
       "3  {'name': 'Zensity v', 'teamId': 11, 'health': ...    None          0.0   \n",
       "4  {'name': 'Anerge', 'teamId': 11, 'health': 100...    None          0.0   \n",
       "\n",
       "   numAlivePlayers         common                        _D  \\\n",
       "0             51.0  {'isGame': 0}  2020-04-26T07:58:25.701Z   \n",
       "1             51.0  {'isGame': 0}  2020-04-26T07:58:25.731Z   \n",
       "2             52.0  {'isGame': 0}  2020-04-26T07:58:26.263Z   \n",
       "3             57.0  {'isGame': 0}  2020-04-26T07:58:27.195Z   \n",
       "4             57.0  {'isGame': 0}  2020-04-26T07:58:27.196Z   \n",
       "\n",
       "                  _T  distance  \\\n",
       "0  LogPlayerPosition       NaN   \n",
       "1  LogPlayerPosition       NaN   \n",
       "2  LogPlayerPosition       NaN   \n",
       "3  LogPlayerPosition       NaN   \n",
       "4  LogPlayerPosition       NaN   \n",
       "\n",
       "                                             MatchId  \n",
       "0  match.bro.official.console-06.console.squad.na...  \n",
       "1  match.bro.official.console-06.console.squad.na...  \n",
       "2  match.bro.official.console-06.console.squad.na...  \n",
       "3  match.bro.official.console-06.console.squad.na...  \n",
       "4  match.bro.official.console-06.console.squad.na...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# copy match id to all rows and drop LogMatchDefinition event\n",
    "# we will use this later to add use it for future merging tasks\n",
    "matchid = json.loads(json_data)[0][\"MatchId\"]\n",
    "data['MatchId'] = matchid\n",
    "\n",
    "#reset indexes\n",
    "data.reset_index(drop=True, inplace=True)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['character', 'vehicle', 'elapsedTime', 'numAlivePlayers', 'common',\n",
       "       '_D', '_T', 'distance', 'MatchId'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extracting Player information variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vehicle</th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>numAlivePlayers</th>\n",
       "      <th>common</th>\n",
       "      <th>_D</th>\n",
       "      <th>_T</th>\n",
       "      <th>distance</th>\n",
       "      <th>MatchId</th>\n",
       "      <th>name</th>\n",
       "      <th>teamId</th>\n",
       "      <th>health</th>\n",
       "      <th>ranking</th>\n",
       "      <th>accountId</th>\n",
       "      <th>isInBlueZone</th>\n",
       "      <th>isInRedZone</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>location.z</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.701Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>heyyzeus_</td>\n",
       "      <td>3</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.2144d6e42697484a9cb897bfa23877aa</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>795399.500000</td>\n",
       "      <td>19790.187500</td>\n",
       "      <td>530.360596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:25.731Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>BERRYWEBBLE3000</td>\n",
       "      <td>10</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.86bec48fc0844e36a449bb7eea5c740d</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>118560.398438</td>\n",
       "      <td>138603.453125</td>\n",
       "      <td>2340.935059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:26.263Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343746.000000</td>\n",
       "      <td>170965.203125</td>\n",
       "      <td>1531.633667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.195Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>Zensity v</td>\n",
       "      <td>11</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.a1b13abbcb7248998806b6ccffb2bad2</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343813.187500</td>\n",
       "      <td>170571.140625</td>\n",
       "      <td>1453.433105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>{'isGame': 0}</td>\n",
       "      <td>2020-04-26T07:58:27.196Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>Anerge</td>\n",
       "      <td>11</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.1c04824e2c3849169e6e06ae1cc67519</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343090.125000</td>\n",
       "      <td>170392.671875</td>\n",
       "      <td>1453.433105</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  vehicle  elapsedTime  numAlivePlayers         common  \\\n",
       "0    None          0.0             51.0  {'isGame': 0}   \n",
       "1    None          0.0             51.0  {'isGame': 0}   \n",
       "2    None          0.0             52.0  {'isGame': 0}   \n",
       "3    None          0.0             57.0  {'isGame': 0}   \n",
       "4    None          0.0             57.0  {'isGame': 0}   \n",
       "\n",
       "                         _D                 _T  distance  \\\n",
       "0  2020-04-26T07:58:25.701Z  LogPlayerPosition       NaN   \n",
       "1  2020-04-26T07:58:25.731Z  LogPlayerPosition       NaN   \n",
       "2  2020-04-26T07:58:26.263Z  LogPlayerPosition       NaN   \n",
       "3  2020-04-26T07:58:27.195Z  LogPlayerPosition       NaN   \n",
       "4  2020-04-26T07:58:27.196Z  LogPlayerPosition       NaN   \n",
       "\n",
       "                                             MatchId             name  teamId  \\\n",
       "0  match.bro.official.console-06.console.squad.na...        heyyzeus_       3   \n",
       "1  match.bro.official.console-06.console.squad.na...  BERRYWEBBLE3000      10   \n",
       "2  match.bro.official.console-06.console.squad.na...         walas-ii       1   \n",
       "3  match.bro.official.console-06.console.squad.na...        Zensity v      11   \n",
       "4  match.bro.official.console-06.console.squad.na...           Anerge      11   \n",
       "\n",
       "   health  ranking                                 accountId  isInBlueZone  \\\n",
       "0   100.0        0  account.2144d6e42697484a9cb897bfa23877aa         False   \n",
       "1   100.0        0  account.86bec48fc0844e36a449bb7eea5c740d         False   \n",
       "2   100.0        0  account.f915e33e51374cb7888f485a91331417         False   \n",
       "3   100.0        0  account.a1b13abbcb7248998806b6ccffb2bad2         False   \n",
       "4   100.0        0  account.1c04824e2c3849169e6e06ae1cc67519         False   \n",
       "\n",
       "   isInRedZone     location.x     location.y   location.z  \n",
       "0        False  795399.500000   19790.187500   530.360596  \n",
       "1        False  118560.398438  138603.453125  2340.935059  \n",
       "2        False  343746.000000  170965.203125  1531.633667  \n",
       "3        False  343813.187500  170571.140625  1453.433105  \n",
       "4        False  343090.125000  170392.671875  1453.433105  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get player information\n",
    "from pandas.io.json import json_normalize\n",
    "# we need to expand the json documents on the character column\n",
    "data = pd.concat([data, json_normalize(data['character'])], axis=1).drop(['character', 'zone'], axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set isGame variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def isGame(x):\n",
    "    return x['isGame']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>vehicle</th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>numAlivePlayers</th>\n",
       "      <th>_D</th>\n",
       "      <th>_T</th>\n",
       "      <th>distance</th>\n",
       "      <th>MatchId</th>\n",
       "      <th>name</th>\n",
       "      <th>teamId</th>\n",
       "      <th>health</th>\n",
       "      <th>ranking</th>\n",
       "      <th>accountId</th>\n",
       "      <th>isInBlueZone</th>\n",
       "      <th>isInRedZone</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>location.z</th>\n",
       "      <th>isGame</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>2020-04-26T07:58:25.701Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>heyyzeus_</td>\n",
       "      <td>3</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.2144d6e42697484a9cb897bfa23877aa</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>795399.500000</td>\n",
       "      <td>19790.187500</td>\n",
       "      <td>530.360596</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>2020-04-26T07:58:25.731Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>BERRYWEBBLE3000</td>\n",
       "      <td>10</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.86bec48fc0844e36a449bb7eea5c740d</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>118560.398438</td>\n",
       "      <td>138603.453125</td>\n",
       "      <td>2340.935059</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2020-04-26T07:58:26.263Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343746.000000</td>\n",
       "      <td>170965.203125</td>\n",
       "      <td>1531.633667</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>2020-04-26T07:58:27.195Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>Zensity v</td>\n",
       "      <td>11</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.a1b13abbcb7248998806b6ccffb2bad2</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343813.187500</td>\n",
       "      <td>170571.140625</td>\n",
       "      <td>1453.433105</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>None</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>2020-04-26T07:58:27.196Z</td>\n",
       "      <td>LogPlayerPosition</td>\n",
       "      <td>NaN</td>\n",
       "      <td>match.bro.official.console-06.console.squad.na...</td>\n",
       "      <td>Anerge</td>\n",
       "      <td>11</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>account.1c04824e2c3849169e6e06ae1cc67519</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>343090.125000</td>\n",
       "      <td>170392.671875</td>\n",
       "      <td>1453.433105</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  vehicle  elapsedTime  numAlivePlayers                        _D  \\\n",
       "0    None          0.0             51.0  2020-04-26T07:58:25.701Z   \n",
       "1    None          0.0             51.0  2020-04-26T07:58:25.731Z   \n",
       "2    None          0.0             52.0  2020-04-26T07:58:26.263Z   \n",
       "3    None          0.0             57.0  2020-04-26T07:58:27.195Z   \n",
       "4    None          0.0             57.0  2020-04-26T07:58:27.196Z   \n",
       "\n",
       "                  _T  distance  \\\n",
       "0  LogPlayerPosition       NaN   \n",
       "1  LogPlayerPosition       NaN   \n",
       "2  LogPlayerPosition       NaN   \n",
       "3  LogPlayerPosition       NaN   \n",
       "4  LogPlayerPosition       NaN   \n",
       "\n",
       "                                             MatchId             name  teamId  \\\n",
       "0  match.bro.official.console-06.console.squad.na...        heyyzeus_       3   \n",
       "1  match.bro.official.console-06.console.squad.na...  BERRYWEBBLE3000      10   \n",
       "2  match.bro.official.console-06.console.squad.na...         walas-ii       1   \n",
       "3  match.bro.official.console-06.console.squad.na...        Zensity v      11   \n",
       "4  match.bro.official.console-06.console.squad.na...           Anerge      11   \n",
       "\n",
       "   health  ranking                                 accountId  isInBlueZone  \\\n",
       "0   100.0        0  account.2144d6e42697484a9cb897bfa23877aa         False   \n",
       "1   100.0        0  account.86bec48fc0844e36a449bb7eea5c740d         False   \n",
       "2   100.0        0  account.f915e33e51374cb7888f485a91331417         False   \n",
       "3   100.0        0  account.a1b13abbcb7248998806b6ccffb2bad2         False   \n",
       "4   100.0        0  account.1c04824e2c3849169e6e06ae1cc67519         False   \n",
       "\n",
       "   isInRedZone     location.x     location.y   location.z  isGame  \n",
       "0        False  795399.500000   19790.187500   530.360596     0.0  \n",
       "1        False  118560.398438  138603.453125  2340.935059     0.0  \n",
       "2        False  343746.000000  170965.203125  1531.633667     0.0  \n",
       "3        False  343813.187500  170571.140625  1453.433105     0.0  \n",
       "4        False  343090.125000  170392.671875  1453.433105     0.0  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['isGame'] = data['common'].apply(isGame)\n",
    "data = data.drop('common', axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we will use the vehicle to check what players are still in the airplane\n",
    "data.vehicle = data.vehicle.apply(lambda x: {} if pd.isna(x) else x)\n",
    "data['vehicle'] = json_normalize(data.vehicle)['vehicleType']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpolate players location for all elapsed times"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0 has 625 observations\n",
      "1.0 has 6 observations\n",
      "2.0 has 12 observations\n",
      "3.0 has 22 observations\n",
      "4.0 has 19 observations\n",
      "5.0 has 6 observations\n",
      "6.0 has 1 observations\n",
      "7.0 has 9 observations\n",
      "8.0 has 12 observations\n",
      "9.0 has 6 observations\n",
      "10.0 has 3 observations\n",
      "11.0 has 7 observations\n",
      "12.0 has 15 observations\n",
      "13.0 has 21 observations\n",
      "14.0 has 19 observations\n",
      "15.0 has 3 observations\n",
      "16.0 has 1 observations\n",
      "17.0 has 11 observations\n",
      "18.0 has 11 observations\n",
      "19.0 has 5 observations\n",
      "20.0 has 4 observations\n",
      "21.0 has 7 observations\n",
      "22.0 has 17 observations\n",
      "23.0 has 20 observations\n",
      "24.0 has 19 observations\n",
      "25.0 has 1 observations\n",
      "26.0 has 1 observations\n",
      "27.0 has 14 observations\n",
      "28.0 has 8 observations\n",
      "29.0 has 5 observations\n",
      "30.0 has 4 observations\n",
      "31.0 has 10 observations\n",
      "32.0 has 15 observations\n",
      "33.0 has 20 observations\n",
      "34.0 has 18 observations\n",
      "35.0 has 1 observations\n",
      "36.0 has 3 observations\n",
      "37.0 has 14 observations\n",
      "38.0 has 6 observations\n",
      "39.0 has 6 observations\n",
      "40.0 has 4 observations\n",
      "41.0 has 12 observations\n",
      "42.0 has 15 observations\n",
      "43.0 has 19 observations\n",
      "44.0 has 16 observations\n",
      "45.0 has 1 observations\n",
      "46.0 has 4 observations\n",
      "47.0 has 14 observations\n",
      "48.0 has 6 observations\n",
      "49.0 has 5 observations\n",
      "50.0 has 4 observations\n",
      "51.0 has 13 observations\n",
      "52.0 has 17 observations\n",
      "53.0 has 20 observations\n",
      "54.0 has 13 observations\n",
      "nan has 0 observations\n",
      "56.0 has 7 observations\n",
      "57.0 has 11 observations\n",
      "58.0 has 8 observations\n",
      "59.0 has 3 observations\n",
      "60.0 has 5 observations\n",
      "61.0 has 12 observations\n",
      "62.0 has 18 observations\n",
      "63.0 has 20 observations\n",
      "64.0 has 12 observations\n",
      "66.0 has 8 observations\n",
      "67.0 has 10 observations\n",
      "68.0 has 8 observations\n",
      "69.0 has 3 observations\n",
      "70.0 has 8 observations\n",
      "71.0 has 10 observations\n",
      "72.0 has 20 observations\n",
      "73.0 has 20 observations\n",
      "74.0 has 9 observations\n",
      "75.0 has 1 observations\n",
      "76.0 has 7 observations\n",
      "77.0 has 12 observations\n",
      "78.0 has 6 observations\n",
      "79.0 has 4 observations\n",
      "80.0 has 7 observations\n",
      "81.0 has 10 observations\n",
      "82.0 has 21 observations\n",
      "83.0 has 22 observations\n",
      "84.0 has 6 observations\n",
      "85.0 has 1 observations\n",
      "86.0 has 7 observations\n",
      "87.0 has 12 observations\n",
      "88.0 has 7 observations\n",
      "89.0 has 4 observations\n",
      "90.0 has 6 observations\n",
      "91.0 has 12 observations\n",
      "92.0 has 22 observations\n",
      "93.0 has 19 observations\n",
      "94.0 has 6 observations\n",
      "95.0 has 1 observations\n",
      "96.0 has 8 observations\n",
      "97.0 has 12 observations\n",
      "98.0 has 6 observations\n",
      "99.0 has 4 observations\n",
      "100.0 has 7 observations\n",
      "101.0 has 15 observations\n",
      "102.0 has 19 observations\n",
      "103.0 has 19 observations\n",
      "104.0 has 5 observations\n",
      "105.0 has 1 observations\n",
      "106.0 has 11 observations\n",
      "107.0 has 10 observations\n",
      "108.0 has 6 observations\n",
      "109.0 has 4 observations\n",
      "110.0 has 7 observations\n",
      "111.0 has 17 observations\n",
      "112.0 has 20 observations\n",
      "113.0 has 19 observations\n",
      "114.0 has 1 observations\n",
      "115.0 has 1 observations\n",
      "116.0 has 16 observations\n",
      "117.0 has 6 observations\n",
      "118.0 has 6 observations\n",
      "119.0 has 3 observations\n",
      "120.0 has 10 observations\n",
      "121.0 has 16 observations\n",
      "122.0 has 21 observations\n",
      "123.0 has 16 observations\n",
      "124.0 has 1 observations\n",
      "125.0 has 4 observations\n",
      "126.0 has 14 observations\n",
      "127.0 has 5 observations\n",
      "128.0 has 6 observations\n",
      "129.0 has 3 observations\n",
      "130.0 has 12 observations\n",
      "131.0 has 15 observations\n",
      "132.0 has 20 observations\n",
      "133.0 has 15 observations\n",
      "134.0 has 1 observations\n",
      "135.0 has 6 observations\n",
      "136.0 has 12 observations\n",
      "137.0 has 7 observations\n",
      "138.0 has 4 observations\n",
      "139.0 has 4 observations\n",
      "140.0 has 12 observations\n",
      "141.0 has 18 observations\n",
      "142.0 has 21 observations\n",
      "143.0 has 10 observations\n",
      "145.0 has 8 observations\n",
      "146.0 has 10 observations\n",
      "147.0 has 8 observations\n",
      "148.0 has 2 observations\n",
      "149.0 has 7 observations\n",
      "150.0 has 10 observations\n",
      "151.0 has 19 observations\n",
      "152.0 has 18 observations\n",
      "153.0 has 7 observations\n",
      "154.0 has 1 observations\n",
      "155.0 has 7 observations\n",
      "156.0 has 12 observations\n",
      "157.0 has 5 observations\n",
      "158.0 has 4 observations\n",
      "159.0 has 2 observations\n",
      "160.0 has 12 observations\n",
      "161.0 has 17 observations\n",
      "162.0 has 17 observations\n",
      "163.0 has 6 observations\n",
      "164.0 has 1 observations\n",
      "165.0 has 7 observations\n",
      "166.0 has 12 observations\n",
      "167.0 has 5 observations\n",
      "168.0 has 4 observations\n",
      "169.0 has 3 observations\n",
      "170.0 has 12 observations\n",
      "171.0 has 17 observations\n",
      "172.0 has 16 observations\n",
      "173.0 has 6 observations\n",
      "174.0 has 1 observations\n",
      "175.0 has 10 observations\n",
      "176.0 has 11 observations\n",
      "177.0 has 4 observations\n",
      "178.0 has 3 observations\n",
      "179.0 has 3 observations\n",
      "180.0 has 15 observations\n",
      "181.0 has 16 observations\n",
      "182.0 has 17 observations\n",
      "183.0 has 3 observations\n",
      "184.0 has 1 observations\n",
      "185.0 has 12 observations\n",
      "186.0 has 10 observations\n",
      "187.0 has 3 observations\n",
      "188.0 has 4 observations\n",
      "189.0 has 4 observations\n",
      "190.0 has 16 observations\n",
      "191.0 has 15 observations\n",
      "192.0 has 17 observations\n",
      "193.0 has 1 observations\n",
      "194.0 has 2 observations\n",
      "195.0 has 14 observations\n",
      "196.0 has 6 observations\n",
      "197.0 has 3 observations\n",
      "198.0 has 3 observations\n",
      "199.0 has 5 observations\n",
      "200.0 has 15 observations\n",
      "201.0 has 15 observations\n",
      "202.0 has 15 observations\n",
      "203.0 has 1 observations\n",
      "204.0 has 3 observations\n",
      "205.0 has 12 observations\n",
      "206.0 has 5 observations\n",
      "207.0 has 3 observations\n",
      "208.0 has 3 observations\n",
      "209.0 has 7 observations\n",
      "210.0 has 17 observations\n",
      "211.0 has 14 observations\n",
      "212.0 has 12 observations\n",
      "214.0 has 6 observations\n",
      "215.0 has 9 observations\n",
      "216.0 has 6 observations\n",
      "217.0 has 1 observations\n",
      "218.0 has 3 observations\n",
      "219.0 has 8 observations\n",
      "220.0 has 16 observations\n",
      "221.0 has 17 observations\n",
      "222.0 has 9 observations\n",
      "224.0 has 8 observations\n",
      "225.0 has 7 observations\n",
      "226.0 has 6 observations\n",
      "227.0 has 1 observations\n",
      "228.0 has 4 observations\n",
      "229.0 has 7 observations\n",
      "230.0 has 18 observations\n",
      "231.0 has 16 observations\n",
      "232.0 has 8 observations\n",
      "233.0 has 1 observations\n",
      "234.0 has 7 observations\n",
      "235.0 has 9 observations\n",
      "236.0 has 4 observations\n",
      "237.0 has 3 observations\n",
      "238.0 has 2 observations\n",
      "239.0 has 9 observations\n",
      "240.0 has 18 observations\n",
      "241.0 has 16 observations\n",
      "242.0 has 6 observations\n",
      "243.0 has 1 observations\n",
      "244.0 has 7 observations\n",
      "245.0 has 9 observations\n",
      "246.0 has 4 observations\n",
      "247.0 has 3 observations\n",
      "248.0 has 2 observations\n",
      "249.0 has 12 observations\n",
      "250.0 has 17 observations\n",
      "251.0 has 15 observations\n",
      "252.0 has 5 observations\n",
      "253.0 has 1 observations\n",
      "254.0 has 9 observations\n",
      "255.0 has 9 observations\n",
      "256.0 has 3 observations\n",
      "257.0 has 2 observations\n",
      "258.0 has 2 observations\n",
      "259.0 has 15 observations\n",
      "260.0 has 14 observations\n",
      "261.0 has 16 observations\n",
      "262.0 has 2 observations\n",
      "263.0 has 1 observations\n",
      "264.0 has 11 observations\n",
      "265.0 has 8 observations\n",
      "266.0 has 2 observations\n",
      "267.0 has 2 observations\n",
      "268.0 has 4 observations\n",
      "269.0 has 15 observations\n",
      "270.0 has 14 observations\n",
      "271.0 has 15 observations\n",
      "272.0 has 1 observations\n",
      "273.0 has 3 observations\n",
      "274.0 has 12 observations\n",
      "275.0 has 5 observations\n",
      "276.0 has 2 observations\n",
      "277.0 has 2 observations\n",
      "278.0 has 7 observations\n",
      "279.0 has 14 observations\n",
      "280.0 has 14 observations\n",
      "281.0 has 13 observations\n",
      "282.0 has 1 observations\n",
      "283.0 has 6 observations\n",
      "284.0 has 9 observations\n",
      "285.0 has 6 observations\n",
      "286.0 has 1 observations\n",
      "287.0 has 2 observations\n",
      "288.0 has 7 observations\n",
      "289.0 has 16 observations\n",
      "290.0 has 15 observations\n",
      "291.0 has 9 observations\n",
      "293.0 has 8 observations\n",
      "294.0 has 7 observations\n",
      "295.0 has 6 observations\n",
      "296.0 has 1 observations\n",
      "297.0 has 3 observations\n",
      "298.0 has 7 observations\n",
      "299.0 has 16 observations\n",
      "300.0 has 13 observations\n",
      "301.0 has 8 observations\n",
      "302.0 has 1 observations\n",
      "303.0 has 7 observations\n",
      "304.0 has 9 observations\n",
      "305.0 has 4 observations\n",
      "306.0 has 2 observations\n",
      "307.0 has 2 observations\n",
      "308.0 has 9 observations\n",
      "309.0 has 15 observations\n",
      "310.0 has 14 observations\n",
      "311.0 has 6 observations\n",
      "312.0 has 1 observations\n",
      "313.0 has 8 observations\n",
      "314.0 has 10 observations\n",
      "315.0 has 3 observations\n",
      "316.0 has 1 observations\n",
      "317.0 has 2 observations\n",
      "318.0 has 13 observations\n",
      "319.0 has 12 observations\n",
      "320.0 has 13 observations\n",
      "321.0 has 5 observations\n",
      "322.0 has 1 observations\n",
      "323.0 has 9 observations\n",
      "324.0 has 9 observations\n",
      "325.0 has 3 observations\n",
      "326.0 has 2 observations\n",
      "327.0 has 2 observations\n",
      "328.0 has 15 observations\n",
      "329.0 has 12 observations\n",
      "330.0 has 14 observations\n",
      "331.0 has 1 observations\n",
      "332.0 has 1 observations\n",
      "333.0 has 11 observations\n",
      "334.0 has 8 observations\n",
      "335.0 has 2 observations\n",
      "336.0 has 2 observations\n",
      "337.0 has 2 observations\n",
      "338.0 has 15 observations\n",
      "339.0 has 12 observations\n",
      "340.0 has 14 observations\n",
      "341.0 has 1 observations\n",
      "342.0 has 3 observations\n",
      "343.0 has 11 observations\n",
      "344.0 has 5 observations\n",
      "345.0 has 2 observations\n",
      "346.0 has 2 observations\n",
      "347.0 has 6 observations\n",
      "348.0 has 12 observations\n",
      "349.0 has 11 observations\n",
      "350.0 has 13 observations\n",
      "351.0 has 1 observations\n",
      "352.0 has 4 observations\n",
      "353.0 has 11 observations\n",
      "354.0 has 4 observations\n",
      "355.0 has 2 observations\n",
      "356.0 has 2 observations\n",
      "357.0 has 7 observations\n",
      "358.0 has 11 observations\n",
      "359.0 has 14 observations\n",
      "360.0 has 9 observations\n",
      "362.0 has 8 observations\n",
      "363.0 has 7 observations\n",
      "364.0 has 5 observations\n",
      "365.0 has 1 observations\n",
      "366.0 has 2 observations\n",
      "367.0 has 7 observations\n",
      "368.0 has 12 observations\n",
      "369.0 has 13 observations\n",
      "370.0 has 8 observations\n",
      "371.0 has 1 observations\n",
      "372.0 has 7 observations\n",
      "373.0 has 9 observations\n",
      "374.0 has 3 observations\n",
      "375.0 has 1 observations\n",
      "376.0 has 2 observations\n",
      "377.0 has 7 observations\n",
      "378.0 has 13 observations\n",
      "379.0 has 14 observations\n",
      "380.0 has 6 observations\n",
      "381.0 has 1 observations\n",
      "382.0 has 7 observations\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "383.0 has 10 observations\n",
      "384.0 has 2 observations\n",
      "385.0 has 2 observations\n",
      "386.0 has 1 observations\n",
      "387.0 has 10 observations\n",
      "388.0 has 12 observations\n",
      "389.0 has 14 observations\n",
      "390.0 has 4 observations\n",
      "391.0 has 1 observations\n",
      "392.0 has 9 observations\n",
      "393.0 has 9 observations\n",
      "394.0 has 2 observations\n",
      "395.0 has 2 observations\n",
      "396.0 has 1 observations\n",
      "397.0 has 12 observations\n",
      "398.0 has 11 observations\n",
      "399.0 has 13 observations\n",
      "400.0 has 3 observations\n",
      "401.0 has 1 observations\n",
      "402.0 has 11 observations\n",
      "403.0 has 7 observations\n",
      "404.0 has 2 observations\n",
      "405.0 has 2 observations\n",
      "406.0 has 3 observations\n",
      "407.0 has 11 observations\n",
      "408.0 has 12 observations\n",
      "409.0 has 13 observations\n",
      "410.0 has 1 observations\n",
      "411.0 has 3 observations\n",
      "412.0 has 12 observations\n",
      "413.0 has 4 observations\n",
      "414.0 has 2 observations\n",
      "415.0 has 2 observations\n",
      "416.0 has 4 observations\n",
      "417.0 has 10 observations\n",
      "418.0 has 11 observations\n",
      "419.0 has 13 observations\n",
      "420.0 has 1 observations\n",
      "421.0 has 5 observations\n",
      "422.0 has 9 observations\n",
      "423.0 has 5 observations\n",
      "424.0 has 1 observations\n",
      "425.0 has 1 observations\n",
      "426.0 has 5 observations\n",
      "427.0 has 11 observations\n",
      "428.0 has 14 observations\n",
      "429.0 has 9 observations\n",
      "431.0 has 8 observations\n",
      "432.0 has 6 observations\n",
      "433.0 has 5 observations\n",
      "434.0 has 1 observations\n",
      "435.0 has 1 observations\n",
      "436.0 has 6 observations\n",
      "437.0 has 12 observations\n",
      "438.0 has 12 observations\n",
      "439.0 has 8 observations\n",
      "440.0 has 1 observations\n",
      "441.0 has 7 observations\n",
      "442.0 has 8 observations\n",
      "443.0 has 3 observations\n",
      "444.0 has 1 observations\n",
      "446.0 has 6 observations\n",
      "447.0 has 11 observations\n",
      "448.0 has 12 observations\n",
      "449.0 has 6 observations\n",
      "450.0 has 1 observations\n",
      "451.0 has 7 observations\n",
      "452.0 has 8 observations\n",
      "453.0 has 3 observations\n",
      "454.0 has 1 observations\n",
      "456.0 has 7 observations\n",
      "457.0 has 10 observations\n",
      "458.0 has 11 observations\n",
      "459.0 has 6 observations\n",
      "460.0 has 1 observations\n",
      "461.0 has 9 observations\n",
      "462.0 has 8 observations\n",
      "463.0 has 2 observations\n",
      "466.0 has 9 observations\n",
      "467.0 has 9 observations\n",
      "468.0 has 13 observations\n",
      "469.0 has 3 observations\n",
      "470.0 has 1 observations\n",
      "471.0 has 10 observations\n",
      "472.0 has 7 observations\n",
      "473.0 has 2 observations\n",
      "476.0 has 9 observations\n",
      "477.0 has 11 observations\n",
      "478.0 has 13 observations\n",
      "479.0 has 1 observations\n",
      "480.0 has 3 observations\n",
      "481.0 has 11 observations\n",
      "482.0 has 4 observations\n",
      "483.0 has 2 observations\n",
      "485.0 has 4 observations\n",
      "486.0 has 7 observations\n",
      "487.0 has 10 observations\n",
      "488.0 has 12 observations\n",
      "489.0 has 1 observations\n",
      "490.0 has 4 observations\n",
      "491.0 has 10 observations\n",
      "492.0 has 4 observations\n",
      "493.0 has 2 observations\n",
      "495.0 has 5 observations\n",
      "496.0 has 8 observations\n",
      "497.0 has 11 observations\n",
      "498.0 has 10 observations\n",
      "500.0 has 8 observations\n",
      "501.0 has 6 observations\n",
      "502.0 has 5 observations\n",
      "503.0 has 1 observations\n",
      "505.0 has 5 observations\n",
      "506.0 has 8 observations\n",
      "507.0 has 13 observations\n",
      "508.0 has 8 observations\n",
      "510.0 has 8 observations\n",
      "511.0 has 8 observations\n",
      "512.0 has 3 observations\n",
      "513.0 has 1 observations\n",
      "515.0 has 6 observations\n",
      "516.0 has 10 observations\n",
      "517.0 has 12 observations\n",
      "518.0 has 6 observations\n",
      "519.0 has 1 observations\n",
      "520.0 has 7 observations\n",
      "521.0 has 8 observations\n",
      "522.0 has 3 observations\n",
      "523.0 has 1 observations\n",
      "525.0 has 7 observations\n",
      "526.0 has 10 observations\n",
      "527.0 has 12 observations\n",
      "528.0 has 5 observations\n",
      "529.0 has 1 observations\n",
      "530.0 has 9 observations\n",
      "531.0 has 8 observations\n",
      "532.0 has 2 observations\n",
      "535.0 has 7 observations\n",
      "536.0 has 11 observations\n",
      "537.0 has 13 observations\n",
      "538.0 has 3 observations\n",
      "539.0 has 1 observations\n",
      "540.0 has 10 observations\n",
      "541.0 has 7 observations\n",
      "542.0 has 2 observations\n",
      "545.0 has 9 observations\n",
      "546.0 has 11 observations\n",
      "547.0 has 13 observations\n",
      "548.0 has 1 observations\n",
      "549.0 has 3 observations\n",
      "550.0 has 11 observations\n",
      "551.0 has 4 observations\n",
      "552.0 has 2 observations\n",
      "554.0 has 2 observations\n",
      "555.0 has 9 observations\n",
      "556.0 has 10 observations\n",
      "557.0 has 12 observations\n",
      "558.0 has 1 observations\n",
      "559.0 has 4 observations\n",
      "560.0 has 10 observations\n",
      "561.0 has 4 observations\n",
      "562.0 has 2 observations\n",
      "564.0 has 4 observations\n",
      "565.0 has 9 observations\n",
      "566.0 has 10 observations\n",
      "567.0 has 11 observations\n",
      "569.0 has 6 observations\n",
      "570.0 has 8 observations\n",
      "571.0 has 5 observations\n",
      "572.0 has 1 observations\n",
      "574.0 has 5 observations\n",
      "575.0 has 8 observations\n",
      "576.0 has 11 observations\n",
      "577.0 has 8 observations\n",
      "579.0 has 8 observations\n",
      "580.0 has 6 observations\n",
      "581.0 has 5 observations\n",
      "582.0 has 1 observations\n",
      "584.0 has 5 observations\n",
      "585.0 has 9 observations\n",
      "586.0 has 10 observations\n",
      "587.0 has 6 observations\n",
      "588.0 has 1 observations\n",
      "589.0 has 7 observations\n",
      "590.0 has 8 observations\n",
      "591.0 has 3 observations\n",
      "592.0 has 1 observations\n",
      "594.0 has 7 observations\n",
      "595.0 has 9 observations\n",
      "596.0 has 9 observations\n",
      "597.0 has 5 observations\n",
      "598.0 has 1 observations\n",
      "599.0 has 7 observations\n",
      "600.0 has 8 observations\n",
      "601.0 has 3 observations\n",
      "602.0 has 1 observations\n",
      "604.0 has 7 observations\n",
      "605.0 has 9 observations\n",
      "606.0 has 9 observations\n",
      "607.0 has 5 observations\n",
      "608.0 has 1 observations\n",
      "609.0 has 8 observations\n",
      "610.0 has 9 observations\n",
      "611.0 has 2 observations\n",
      "614.0 has 7 observations\n",
      "615.0 has 10 observations\n",
      "616.0 has 11 observations\n",
      "617.0 has 2 observations\n",
      "618.0 has 1 observations\n",
      "619.0 has 11 observations\n",
      "620.0 has 6 observations\n",
      "621.0 has 2 observations\n",
      "623.0 has 2 observations\n",
      "624.0 has 9 observations\n",
      "625.0 has 7 observations\n",
      "626.0 has 11 observations\n",
      "627.0 has 1 observations\n",
      "628.0 has 4 observations\n",
      "629.0 has 10 observations\n",
      "630.0 has 4 observations\n",
      "631.0 has 2 observations\n",
      "633.0 has 5 observations\n",
      "634.0 has 8 observations\n",
      "635.0 has 7 observations\n",
      "636.0 has 10 observations\n",
      "638.0 has 6 observations\n",
      "639.0 has 8 observations\n",
      "640.0 has 5 observations\n",
      "641.0 has 1 observations\n",
      "643.0 has 5 observations\n",
      "644.0 has 8 observations\n",
      "645.0 has 11 observations\n",
      "646.0 has 6 observations\n",
      "647.0 has 1 observations\n",
      "648.0 has 7 observations\n",
      "649.0 has 6 observations\n",
      "650.0 has 5 observations\n",
      "651.0 has 1 observations\n",
      "653.0 has 6 observations\n",
      "654.0 has 10 observations\n",
      "655.0 has 9 observations\n",
      "656.0 has 5 observations\n",
      "657.0 has 1 observations\n",
      "658.0 has 8 observations\n",
      "659.0 has 8 observations\n",
      "660.0 has 2 observations\n",
      "661.0 has 1 observations\n",
      "663.0 has 7 observations\n",
      "664.0 has 9 observations\n",
      "665.0 has 9 observations\n",
      "666.0 has 4 observations\n",
      "667.0 has 1 observations\n",
      "668.0 has 9 observations\n",
      "669.0 has 8 observations\n",
      "670.0 has 2 observations\n",
      "673.0 has 7 observations\n",
      "674.0 has 10 observations\n",
      "675.0 has 9 observations\n",
      "676.0 has 3 observations\n",
      "677.0 has 1 observations\n",
      "678.0 has 9 observations\n",
      "679.0 has 8 observations\n",
      "680.0 has 2 observations\n",
      "683.0 has 9 observations\n",
      "684.0 has 9 observations\n",
      "685.0 has 10 observations\n",
      "686.0 has 1 observations\n",
      "687.0 has 1 observations\n",
      "688.0 has 11 observations\n",
      "689.0 has 6 observations\n",
      "690.0 has 2 observations\n",
      "692.0 has 2 observations\n",
      "693.0 has 8 observations\n",
      "694.0 has 8 observations\n",
      "695.0 has 10 observations\n",
      "696.0 has 1 observations\n",
      "697.0 has 3 observations\n",
      "698.0 has 11 observations\n",
      "699.0 has 4 observations\n",
      "700.0 has 2 observations\n",
      "702.0 has 4 observations\n",
      "703.0 has 6 observations\n",
      "704.0 has 7 observations\n",
      "705.0 has 10 observations\n",
      "706.0 has 1 observations\n",
      "707.0 has 4 observations\n",
      "708.0 has 10 observations\n",
      "709.0 has 4 observations\n",
      "710.0 has 2 observations\n",
      "712.0 has 5 observations\n",
      "713.0 has 7 observations\n",
      "714.0 has 7 observations\n",
      "715.0 has 7 observations\n",
      "717.0 has 8 observations\n",
      "718.0 has 6 observations\n",
      "719.0 has 5 observations\n",
      "720.0 has 1 observations\n",
      "722.0 has 5 observations\n",
      "723.0 has 8 observations\n",
      "724.0 has 7 observations\n",
      "725.0 has 6 observations\n",
      "726.0 has 1 observations\n",
      "727.0 has 7 observations\n",
      "728.0 has 8 observations\n",
      "729.0 has 3 observations\n",
      "730.0 has 1 observations\n",
      "732.0 has 7 observations\n",
      "733.0 has 8 observations\n",
      "734.0 has 6 observations\n",
      "735.0 has 5 observations\n",
      "736.0 has 1 observations\n",
      "737.0 has 7 observations\n",
      "738.0 has 8 observations\n",
      "739.0 has 3 observations\n",
      "740.0 has 1 observations\n",
      "742.0 has 7 observations\n",
      "743.0 has 8 observations\n",
      "744.0 has 6 observations\n",
      "745.0 has 5 observations\n",
      "746.0 has 1 observations\n",
      "747.0 has 8 observations\n",
      "748.0 has 8 observations\n",
      "749.0 has 2 observations\n",
      "750.0 has 1 observations\n",
      "752.0 has 7 observations\n",
      "753.0 has 8 observations\n",
      "754.0 has 7 observations\n",
      "755.0 has 4 observations\n",
      "756.0 has 1 observations\n",
      "757.0 has 9 observations\n",
      "758.0 has 8 observations\n",
      "759.0 has 2 observations\n",
      "762.0 has 8 observations\n",
      "763.0 has 8 observations\n",
      "764.0 has 9 observations\n",
      "765.0 has 1 observations\n",
      "766.0 has 2 observations\n",
      "767.0 has 12 observations\n",
      "768.0 has 4 observations\n",
      "769.0 has 2 observations\n",
      "771.0 has 2 observations\n",
      "772.0 has 7 observations\n",
      "773.0 has 8 observations\n",
      "774.0 has 8 observations\n",
      "775.0 has 1 observations\n",
      "776.0 has 3 observations\n",
      "777.0 has 11 observations\n",
      "778.0 has 4 observations\n",
      "779.0 has 2 observations\n",
      "781.0 has 4 observations\n",
      "782.0 has 6 observations\n",
      "783.0 has 7 observations\n",
      "784.0 has 8 observations\n",
      "785.0 has 1 observations\n",
      "786.0 has 4 observations\n",
      "787.0 has 10 observations\n",
      "788.0 has 4 observations\n",
      "789.0 has 2 observations\n",
      "791.0 has 5 observations\n",
      "792.0 has 7 observations\n",
      "793.0 has 7 observations\n",
      "794.0 has 7 observations\n",
      "796.0 has 7 observations\n",
      "797.0 has 7 observations\n",
      "798.0 has 5 observations\n",
      "799.0 has 1 observations\n",
      "801.0 has 6 observations\n",
      "802.0 has 8 observations\n",
      "803.0 has 5 observations\n",
      "804.0 has 6 observations\n",
      "805.0 has 1 observations\n",
      "806.0 has 7 observations\n",
      "807.0 has 8 observations\n",
      "808.0 has 3 observations\n",
      "809.0 has 1 observations\n",
      "811.0 has 6 observations\n",
      "812.0 has 8 observations\n",
      "813.0 has 6 observations\n",
      "814.0 has 5 observations\n",
      "815.0 has 1 observations\n",
      "816.0 has 7 observations\n",
      "817.0 has 8 observations\n",
      "818.0 has 3 observations\n",
      "819.0 has 1 observations\n",
      "821.0 has 7 observations\n",
      "822.0 has 7 observations\n",
      "823.0 has 6 observations\n",
      "824.0 has 5 observations\n",
      "825.0 has 1 observations\n",
      "826.0 has 9 observations\n",
      "827.0 has 8 observations\n",
      "828.0 has 2 observations\n",
      "831.0 has 8 observations\n",
      "832.0 has 7 observations\n",
      "833.0 has 8 observations\n",
      "834.0 has 2 observations\n",
      "835.0 has 1 observations\n",
      "836.0 has 9 observations\n",
      "837.0 has 8 observations\n",
      "838.0 has 2 observations\n",
      "841.0 has 7 observations\n",
      "842.0 has 8 observations\n",
      "843.0 has 8 observations\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "844.0 has 1 observations\n",
      "845.0 has 3 observations\n",
      "846.0 has 11 observations\n",
      "847.0 has 4 observations\n",
      "848.0 has 2 observations\n",
      "850.0 has 3 observations\n",
      "851.0 has 6 observations\n",
      "852.0 has 6 observations\n",
      "853.0 has 6 observations\n",
      "854.0 has 1 observations\n",
      "855.0 has 6 observations\n",
      "856.0 has 8 observations\n",
      "857.0 has 4 observations\n",
      "858.0 has 2 observations\n",
      "860.0 has 3 observations\n",
      "861.0 has 7 observations\n",
      "862.0 has 5 observations\n",
      "863.0 has 6 observations\n",
      "865.0 has 8 observations\n",
      "866.0 has 6 observations\n",
      "867.0 has 5 observations\n",
      "868.0 has 1 observations\n",
      "870.0 has 3 observations\n",
      "871.0 has 7 observations\n",
      "872.0 has 5 observations\n",
      "873.0 has 6 observations\n",
      "875.0 has 8 observations\n",
      "876.0 has 6 observations\n",
      "877.0 has 5 observations\n",
      "878.0 has 1 observations\n",
      "880.0 has 4 observations\n",
      "881.0 has 8 observations\n",
      "882.0 has 4 observations\n",
      "883.0 has 5 observations\n",
      "884.0 has 1 observations\n",
      "885.0 has 7 observations\n",
      "886.0 has 8 observations\n",
      "887.0 has 3 observations\n",
      "888.0 has 1 observations\n",
      "890.0 has 5 observations\n",
      "891.0 has 7 observations\n",
      "892.0 has 4 observations\n",
      "893.0 has 5 observations\n",
      "894.0 has 1 observations\n",
      "895.0 has 9 observations\n",
      "896.0 has 8 observations\n",
      "897.0 has 2 observations\n",
      "900.0 has 6 observations\n",
      "901.0 has 7 observations\n",
      "902.0 has 7 observations\n",
      "903.0 has 1 observations\n",
      "904.0 has 3 observations\n",
      "905.0 has 11 observations\n",
      "906.0 has 3 observations\n",
      "907.0 has 2 observations\n",
      "909.0 has 1 observations\n",
      "910.0 has 7 observations\n",
      "911.0 has 6 observations\n",
      "912.0 has 6 observations\n",
      "913.0 has 1 observations\n",
      "914.0 has 5 observations\n",
      "915.0 has 9 observations\n",
      "916.0 has 3 observations\n",
      "917.0 has 2 observations\n",
      "919.0 has 3 observations\n",
      "920.0 has 7 observations\n",
      "921.0 has 5 observations\n",
      "922.0 has 6 observations\n",
      "924.0 has 6 observations\n",
      "925.0 has 8 observations\n",
      "926.0 has 4 observations\n",
      "927.0 has 1 observations\n",
      "929.0 has 3 observations\n",
      "930.0 has 7 observations\n",
      "931.0 has 5 observations\n",
      "932.0 has 6 observations\n",
      "933.0 has 1 observations\n",
      "934.0 has 7 observations\n",
      "935.0 has 6 observations\n",
      "936.0 has 4 observations\n",
      "937.0 has 1 observations\n",
      "939.0 has 4 observations\n",
      "940.0 has 7 observations\n",
      "941.0 has 4 observations\n",
      "942.0 has 6 observations\n",
      "943.0 has 1 observations\n",
      "944.0 has 7 observations\n",
      "945.0 has 7 observations\n",
      "946.0 has 3 observations\n",
      "947.0 has 1 observations\n",
      "949.0 has 5 observations\n",
      "950.0 has 7 observations\n",
      "951.0 has 4 observations\n",
      "952.0 has 5 observations\n",
      "953.0 has 1 observations\n",
      "954.0 has 8 observations\n",
      "955.0 has 8 observations\n",
      "956.0 has 1 observations\n",
      "957.0 has 1 observations\n",
      "959.0 has 5 observations\n",
      "960.0 has 7 observations\n",
      "961.0 has 5 observations\n",
      "962.0 has 4 observations\n",
      "963.0 has 1 observations\n",
      "964.0 has 8 observations\n",
      "965.0 has 7 observations\n",
      "966.0 has 2 observations\n",
      "969.0 has 6 observations\n",
      "970.0 has 7 observations\n",
      "971.0 has 7 observations\n",
      "972.0 has 1 observations\n",
      "973.0 has 2 observations\n",
      "974.0 has 10 observations\n",
      "975.0 has 3 observations\n",
      "976.0 has 2 observations\n",
      "979.0 has 8 observations\n",
      "980.0 has 6 observations\n",
      "981.0 has 6 observations\n",
      "982.0 has 1 observations\n",
      "983.0 has 2 observations\n",
      "984.0 has 8 observations\n",
      "985.0 has 3 observations\n",
      "986.0 has 2 observations\n",
      "988.0 has 2 observations\n",
      "989.0 has 6 observations\n",
      "990.0 has 6 observations\n",
      "991.0 has 6 observations\n",
      "992.0 has 1 observations\n",
      "993.0 has 4 observations\n",
      "994.0 has 6 observations\n",
      "995.0 has 3 observations\n",
      "996.0 has 2 observations\n",
      "998.0 has 3 observations\n",
      "999.0 has 7 observations\n",
      "1000.0 has 5 observations\n",
      "1001.0 has 6 observations\n",
      "1003.0 has 4 observations\n",
      "1004.0 has 6 observations\n",
      "1005.0 has 4 observations\n",
      "1006.0 has 1 observations\n",
      "1008.0 has 3 observations\n",
      "1009.0 has 8 observations\n",
      "1010.0 has 4 observations\n",
      "1011.0 has 6 observations\n",
      "1012.0 has 1 observations\n",
      "1013.0 has 3 observations\n",
      "1014.0 has 7 observations\n",
      "1015.0 has 3 observations\n",
      "1016.0 has 1 observations\n",
      "1018.0 has 5 observations\n",
      "1019.0 has 7 observations\n",
      "1020.0 has 4 observations\n",
      "1021.0 has 5 observations\n",
      "1022.0 has 1 observations\n",
      "1023.0 has 5 observations\n",
      "1024.0 has 7 observations\n",
      "1025.0 has 2 observations\n",
      "1028.0 has 6 observations\n",
      "1029.0 has 7 observations\n",
      "1030.0 has 7 observations\n",
      "1031.0 has 1 observations\n",
      "1032.0 has 1 observations\n",
      "1033.0 has 7 observations\n",
      "1034.0 has 5 observations\n",
      "1035.0 has 2 observations\n",
      "1038.0 has 7 observations\n",
      "1039.0 has 7 observations\n",
      "1040.0 has 6 observations\n",
      "1041.0 has 1 observations\n",
      "1042.0 has 2 observations\n",
      "1043.0 has 8 observations\n",
      "1044.0 has 3 observations\n",
      "1045.0 has 2 observations\n",
      "1047.0 has 3 observations\n",
      "1048.0 has 7 observations\n",
      "1049.0 has 4 observations\n",
      "1050.0 has 7 observations\n",
      "1052.0 has 3 observations\n",
      "1053.0 has 6 observations\n",
      "1054.0 has 4 observations\n",
      "1057.0 has 3 observations\n",
      "1058.0 has 7 observations\n",
      "1059.0 has 5 observations\n",
      "1060.0 has 6 observations\n",
      "1062.0 has 2 observations\n",
      "1063.0 has 6 observations\n",
      "1064.0 has 4 observations\n",
      "1067.0 has 3 observations\n",
      "1068.0 has 7 observations\n",
      "1069.0 has 5 observations\n",
      "1070.0 has 6 observations\n",
      "1071.0 has 1 observations\n",
      "1072.0 has 1 observations\n",
      "1073.0 has 6 observations\n",
      "1074.0 has 4 observations\n",
      "1077.0 has 4 observations\n",
      "1078.0 has 8 observations\n",
      "1079.0 has 4 observations\n",
      "1080.0 has 5 observations\n",
      "1081.0 has 1 observations\n",
      "1082.0 has 1 observations\n",
      "1083.0 has 4 observations\n",
      "1084.0 has 3 observations\n",
      "1087.0 has 5 observations\n",
      "1088.0 has 7 observations\n",
      "1089.0 has 5 observations\n",
      "1090.0 has 4 observations\n",
      "1091.0 has 1 observations\n",
      "1092.0 has 2 observations\n",
      "1093.0 has 5 observations\n",
      "1094.0 has 1 observations\n",
      "1097.0 has 3 observations\n",
      "1098.0 has 7 observations\n",
      "1099.0 has 6 observations\n",
      "1100.0 has 2 observations\n",
      "1101.0 has 1 observations\n",
      "1102.0 has 2 observations\n",
      "1103.0 has 5 observations\n",
      "1104.0 has 1 observations\n",
      "1107.0 has 4 observations\n",
      "1108.0 has 7 observations\n",
      "1109.0 has 6 observations\n",
      "1110.0 has 1 observations\n",
      "1111.0 has 1 observations\n",
      "1112.0 has 4 observations\n",
      "1113.0 has 3 observations\n",
      "1114.0 has 1 observations\n",
      "1117.0 has 5 observations\n",
      "1118.0 has 6 observations\n",
      "1119.0 has 6 observations\n",
      "1120.0 has 1 observations\n",
      "1121.0 has 1 observations\n",
      "1122.0 has 4 observations\n",
      "1123.0 has 3 observations\n",
      "1124.0 has 1 observations\n",
      "1126.0 has 2 observations\n",
      "1127.0 has 3 observations\n",
      "1128.0 has 6 observations\n",
      "1129.0 has 6 observations\n",
      "1130.0 has 1 observations\n",
      "1131.0 has 2 observations\n",
      "1132.0 has 3 observations\n",
      "1133.0 has 4 observations\n",
      "1136.0 has 3 observations\n",
      "1137.0 has 4 observations\n",
      "1138.0 has 5 observations\n",
      "1139.0 has 6 observations\n",
      "1141.0 has 2 observations\n",
      "1142.0 has 3 observations\n",
      "1143.0 has 4 observations\n",
      "1146.0 has 3 observations\n",
      "1147.0 has 5 observations\n",
      "1148.0 has 4 observations\n",
      "1149.0 has 6 observations\n",
      "1150.0 has 1 observations\n",
      "1151.0 has 1 observations\n",
      "1152.0 has 4 observations\n",
      "1153.0 has 3 observations\n",
      "1156.0 has 3 observations\n",
      "1157.0 has 6 observations\n",
      "1158.0 has 4 observations\n",
      "1159.0 has 5 observations\n",
      "1160.0 has 1 observations\n",
      "1161.0 has 2 observations\n",
      "1162.0 has 4 observations\n",
      "1163.0 has 2 observations\n",
      "1166.0 has 3 observations\n",
      "1167.0 has 6 observations\n",
      "1168.0 has 6 observations\n",
      "1169.0 has 3 observations\n",
      "1170.0 has 1 observations\n",
      "1171.0 has 2 observations\n",
      "1172.0 has 5 observations\n",
      "1173.0 has 1 observations\n",
      "1176.0 has 4 observations\n",
      "1177.0 has 7 observations\n",
      "1178.0 has 6 observations\n",
      "1179.0 has 1 observations\n",
      "1180.0 has 1 observations\n",
      "1181.0 has 4 observations\n",
      "1182.0 has 3 observations\n",
      "1183.0 has 1 observations\n",
      "1186.0 has 5 observations\n",
      "1187.0 has 6 observations\n",
      "1188.0 has 6 observations\n",
      "1189.0 has 1 observations\n",
      "1190.0 has 1 observations\n",
      "1191.0 has 4 observations\n",
      "1192.0 has 3 observations\n",
      "1193.0 has 1 observations\n",
      "1195.0 has 3 observations\n",
      "1196.0 has 4 observations\n",
      "1197.0 has 4 observations\n",
      "1198.0 has 7 observations\n",
      "1200.0 has 2 observations\n",
      "1201.0 has 3 observations\n",
      "1202.0 has 4 observations\n",
      "1205.0 has 3 observations\n",
      "1206.0 has 4 observations\n",
      "1207.0 has 5 observations\n",
      "1208.0 has 6 observations\n",
      "1210.0 has 2 observations\n",
      "1211.0 has 3 observations\n",
      "1212.0 has 4 observations\n",
      "1215.0 has 3 observations\n",
      "1216.0 has 5 observations\n",
      "1217.0 has 4 observations\n",
      "1218.0 has 6 observations\n",
      "1219.0 has 1 observations\n",
      "1220.0 has 1 observations\n",
      "1221.0 has 4 observations\n",
      "1222.0 has 3 observations\n",
      "1225.0 has 3 observations\n",
      "1226.0 has 6 observations\n",
      "1227.0 has 4 observations\n",
      "1228.0 has 5 observations\n",
      "1229.0 has 1 observations\n",
      "1230.0 has 1 observations\n",
      "1231.0 has 4 observations\n",
      "1232.0 has 3 observations\n",
      "1235.0 has 3 observations\n",
      "1236.0 has 6 observations\n",
      "1237.0 has 4 observations\n",
      "1238.0 has 5 observations\n",
      "1239.0 has 1 observations\n",
      "1240.0 has 1 observations\n",
      "1241.0 has 6 observations\n",
      "1242.0 has 1 observations\n",
      "1245.0 has 3 observations\n",
      "1246.0 has 6 observations\n",
      "1247.0 has 5 observations\n",
      "1248.0 has 4 observations\n",
      "1249.0 has 1 observations\n",
      "1250.0 has 2 observations\n",
      "1251.0 has 5 observations\n",
      "1252.0 has 1 observations\n",
      "1255.0 has 3 observations\n",
      "1256.0 has 7 observations\n",
      "1257.0 has 5 observations\n",
      "1258.0 has 3 observations\n",
      "1259.0 has 1 observations\n",
      "1260.0 has 2 observations\n",
      "1261.0 has 5 observations\n",
      "1262.0 has 1 observations\n",
      "1265.0 has 4 observations\n",
      "1266.0 has 7 observations\n",
      "1267.0 has 6 observations\n",
      "1268.0 has 1 observations\n",
      "1269.0 has 1 observations\n",
      "1270.0 has 4 observations\n",
      "1271.0 has 1 observations\n",
      "1272.0 has 1 observations\n",
      "1275.0 has 4 observations\n",
      "1276.0 has 7 observations\n",
      "1277.0 has 6 observations\n",
      "1278.0 has 1 observations\n",
      "1279.0 has 1 observations\n",
      "1280.0 has 3 observations\n",
      "1281.0 has 1 observations\n",
      "1284.0 has 2 observations\n",
      "1285.0 has 3 observations\n",
      "1286.0 has 6 observations\n",
      "1287.0 has 6 observations\n",
      "1288.0 has 1 observations\n",
      "1289.0 has 2 observations\n",
      "1290.0 has 2 observations\n",
      "1291.0 has 1 observations\n",
      "1294.0 has 3 observations\n",
      "1295.0 has 4 observations\n",
      "1296.0 has 5 observations\n",
      "1297.0 has 6 observations\n",
      "1299.0 has 2 observations\n",
      "1300.0 has 2 observations\n",
      "1301.0 has 1 observations\n",
      "1304.0 has 3 observations\n",
      "1305.0 has 4 observations\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1306.0 has 5 observations\n",
      "1307.0 has 6 observations\n",
      "1308.0 has 1 observations\n",
      "1309.0 has 1 observations\n",
      "1310.0 has 3 observations\n",
      "1314.0 has 3 observations\n",
      "1315.0 has 6 observations\n",
      "1316.0 has 4 observations\n",
      "1317.0 has 5 observations\n",
      "1318.0 has 1 observations\n",
      "1319.0 has 1 observations\n",
      "1320.0 has 3 observations\n",
      "1324.0 has 3 observations\n",
      "1325.0 has 6 observations\n",
      "1326.0 has 5 observations\n",
      "1327.0 has 4 observations\n",
      "1328.0 has 1 observations\n",
      "1329.0 has 2 observations\n",
      "1330.0 has 2 observations\n",
      "1334.0 has 3 observations\n",
      "1335.0 has 7 observations\n",
      "1336.0 has 6 observations\n",
      "1337.0 has 2 observations\n",
      "1338.0 has 1 observations\n",
      "1339.0 has 2 observations\n",
      "1340.0 has 2 observations\n",
      "1344.0 has 4 observations\n",
      "1345.0 has 6 observations\n",
      "1346.0 has 6 observations\n",
      "1347.0 has 1 observations\n",
      "1348.0 has 1 observations\n",
      "1349.0 has 3 observations\n",
      "1350.0 has 1 observations\n",
      "1354.0 has 4 observations\n",
      "1355.0 has 6 observations\n",
      "1356.0 has 6 observations\n",
      "1357.0 has 1 observations\n",
      "1358.0 has 1 observations\n",
      "1359.0 has 3 observations\n",
      "1360.0 has 1 observations\n",
      "1363.0 has 2 observations\n",
      "1364.0 has 4 observations\n",
      "1365.0 has 4 observations\n",
      "1366.0 has 6 observations\n",
      "1367.0 has 1 observations\n",
      "1368.0 has 2 observations\n",
      "1369.0 has 2 observations\n",
      "1370.0 has 1 observations\n",
      "1373.0 has 3 observations\n",
      "1374.0 has 3 observations\n",
      "1375.0 has 4 observations\n",
      "1376.0 has 7 observations\n",
      "1378.0 has 2 observations\n",
      "1379.0 has 2 observations\n",
      "1380.0 has 1 observations\n",
      "1383.0 has 3 observations\n",
      "1384.0 has 3 observations\n",
      "1385.0 has 5 observations\n",
      "1386.0 has 6 observations\n",
      "1387.0 has 1 observations\n",
      "1388.0 has 1 observations\n",
      "1389.0 has 2 observations\n",
      "1390.0 has 1 observations\n",
      "1393.0 has 3 observations\n",
      "1394.0 has 4 observations\n",
      "1395.0 has 4 observations\n",
      "1396.0 has 6 observations\n",
      "1397.0 has 1 observations\n",
      "1398.0 has 1 observations\n",
      "1399.0 has 3 observations\n",
      "1403.0 has 3 observations\n",
      "1404.0 has 5 observations\n",
      "1405.0 has 4 observations\n",
      "1406.0 has 5 observations\n",
      "1407.0 has 1 observations\n",
      "1408.0 has 2 observations\n",
      "1409.0 has 2 observations\n",
      "1413.0 has 4 observations\n",
      "1414.0 has 5 observations\n",
      "1415.0 has 7 observations\n",
      "1416.0 has 1 observations\n",
      "1417.0 has 1 observations\n",
      "1418.0 has 3 observations\n",
      "1419.0 has 1 observations\n",
      "1423.0 has 4 observations\n",
      "1424.0 has 6 observations\n",
      "1425.0 has 6 observations\n",
      "1426.0 has 1 observations\n",
      "1427.0 has 1 observations\n",
      "1428.0 has 3 observations\n",
      "1429.0 has 1 observations\n",
      "1432.0 has 2 observations\n",
      "1433.0 has 3 observations\n",
      "1434.0 has 5 observations\n",
      "1435.0 has 6 observations\n",
      "1436.0 has 1 observations\n",
      "1437.0 has 2 observations\n",
      "1438.0 has 2 observations\n",
      "1439.0 has 1 observations\n",
      "1442.0 has 2 observations\n",
      "1443.0 has 3 observations\n",
      "1444.0 has 5 observations\n",
      "1445.0 has 6 observations\n",
      "1447.0 has 2 observations\n",
      "1448.0 has 2 observations\n",
      "1449.0 has 1 observations\n",
      "1452.0 has 1 observations\n",
      "1453.0 has 3 observations\n",
      "1454.0 has 4 observations\n",
      "1455.0 has 6 observations\n",
      "1456.0 has 1 observations\n",
      "1457.0 has 1 observations\n",
      "1458.0 has 3 observations\n",
      "1462.0 has 1 observations\n",
      "1463.0 has 5 observations\n",
      "1464.0 has 3 observations\n",
      "1465.0 has 4 observations\n",
      "1466.0 has 1 observations\n",
      "1467.0 has 1 observations\n",
      "1468.0 has 3 observations\n",
      "1472.0 has 1 observations\n",
      "1473.0 has 5 observations\n",
      "1474.0 has 2 observations\n",
      "1475.0 has 4 observations\n",
      "1476.0 has 1 observations\n",
      "1477.0 has 2 observations\n",
      "1478.0 has 2 observations\n",
      "1482.0 has 1 observations\n",
      "1483.0 has 3 observations\n",
      "1484.0 has 4 observations\n",
      "1485.0 has 2 observations\n",
      "1486.0 has 1 observations\n",
      "1487.0 has 2 observations\n",
      "1488.0 has 2 observations\n",
      "1492.0 has 2 observations\n",
      "1493.0 has 3 observations\n",
      "1494.0 has 2 observations\n",
      "1495.0 has 1 observations\n",
      "1496.0 has 1 observations\n",
      "1497.0 has 3 observations\n",
      "1498.0 has 1 observations\n",
      "1502.0 has 2 observations\n",
      "1503.0 has 3 observations\n",
      "1504.0 has 2 observations\n",
      "1505.0 has 1 observations\n",
      "1506.0 has 2 observations\n",
      "1507.0 has 2 observations\n",
      "1508.0 has 1 observations\n",
      "1511.0 has 1 observations\n",
      "1512.0 has 2 observations\n",
      "1513.0 has 2 observations\n",
      "1514.0 has 2 observations\n",
      "1515.0 has 1 observations\n",
      "1516.0 has 2 observations\n",
      "1517.0 has 2 observations\n",
      "1518.0 has 1 observations\n",
      "1521.0 has 1 observations\n",
      "1522.0 has 2 observations\n",
      "1523.0 has 2 observations\n",
      "1524.0 has 3 observations\n",
      "1526.0 has 2 observations\n",
      "1527.0 has 2 observations\n",
      "1528.0 has 1 observations\n",
      "1531.0 has 1 observations\n",
      "1532.0 has 3 observations\n",
      "1533.0 has 1 observations\n",
      "1534.0 has 3 observations\n",
      "1535.0 has 1 observations\n",
      "1536.0 has 1 observations\n",
      "1537.0 has 3 observations\n",
      "1541.0 has 1 observations\n",
      "1542.0 has 3 observations\n",
      "1543.0 has 1 observations\n",
      "1544.0 has 3 observations\n",
      "1545.0 has 1 observations\n",
      "1546.0 has 2 observations\n",
      "1547.0 has 2 observations\n",
      "1551.0 has 2 observations\n",
      "1552.0 has 2 observations\n",
      "1553.0 has 3 observations\n",
      "1554.0 has 1 observations\n",
      "1555.0 has 1 observations\n",
      "1556.0 has 3 observations\n",
      "1557.0 has 1 observations\n",
      "1561.0 has 2 observations\n",
      "1562.0 has 2 observations\n",
      "1563.0 has 2 observations\n",
      "1564.0 has 1 observations\n",
      "1565.0 has 1 observations\n",
      "1566.0 has 3 observations\n",
      "1567.0 has 1 observations\n",
      "1570.0 has 1 observations\n",
      "1571.0 has 1 observations\n",
      "1572.0 has 2 observations\n",
      "1573.0 has 2 observations\n",
      "1574.0 has 1 observations\n",
      "1575.0 has 2 observations\n",
      "1576.0 has 2 observations\n",
      "1577.0 has 1 observations\n",
      "1580.0 has 1 observations\n",
      "1581.0 has 2 observations\n",
      "1582.0 has 1 observations\n",
      "1583.0 has 3 observations\n",
      "1585.0 has 2 observations\n",
      "1586.0 has 2 observations\n",
      "1587.0 has 1 observations\n",
      "1590.0 has 1 observations\n",
      "1591.0 has 2 observations\n",
      "1592.0 has 1 observations\n",
      "1593.0 has 3 observations\n",
      "1594.0 has 1 observations\n",
      "1595.0 has 1 observations\n",
      "1596.0 has 3 observations\n",
      "1601.0 has 2 observations\n",
      "1602.0 has 1 observations\n",
      "1603.0 has 3 observations\n",
      "1604.0 has 1 observations\n",
      "1605.0 has 2 observations\n",
      "1606.0 has 2 observations\n",
      "1611.0 has 2 observations\n",
      "1612.0 has 2 observations\n",
      "1613.0 has 2 observations\n",
      "1614.0 has 1 observations\n",
      "1615.0 has 3 observations\n",
      "1616.0 has 1 observations\n",
      "1620.0 has 1 observations\n",
      "1621.0 has 2 observations\n",
      "1622.0 has 2 observations\n",
      "1623.0 has 1 observations\n",
      "1624.0 has 1 observations\n",
      "1625.0 has 3 observations\n",
      "1626.0 has 1 observations\n",
      "1630.0 has 1 observations\n",
      "1631.0 has 2 observations\n",
      "1632.0 has 2 observations\n",
      "1633.0 has 1 observations\n",
      "1634.0 has 2 observations\n",
      "1635.0 has 2 observations\n",
      "1636.0 has 1 observations\n",
      "1640.0 has 2 observations\n",
      "1641.0 has 1 observations\n",
      "1642.0 has 3 observations\n",
      "1644.0 has 2 observations\n",
      "1645.0 has 2 observations\n",
      "1646.0 has 1 observations\n",
      "1650.0 has 2 observations\n",
      "1651.0 has 1 observations\n",
      "1652.0 has 3 observations\n",
      "1653.0 has 1 observations\n",
      "1654.0 has 1 observations\n",
      "1655.0 has 3 observations\n",
      "1660.0 has 2 observations\n",
      "1661.0 has 1 observations\n",
      "1662.0 has 3 observations\n",
      "1663.0 has 1 observations\n",
      "1664.0 has 1 observations\n",
      "1665.0 has 3 observations\n",
      "1670.0 has 2 observations\n",
      "1671.0 has 1 observations\n",
      "1672.0 has 3 observations\n",
      "1673.0 has 1 observations\n",
      "1674.0 has 2 observations\n",
      "1675.0 has 2 observations\n",
      "1679.0 has 1 observations\n",
      "1680.0 has 1 observations\n",
      "1681.0 has 3 observations\n",
      "1682.0 has 1 observations\n",
      "1683.0 has 1 observations\n",
      "1684.0 has 3 observations\n",
      "1685.0 has 1 observations\n",
      "1689.0 has 1 observations\n",
      "1690.0 has 2 observations\n",
      "1691.0 has 2 observations\n",
      "1692.0 has 1 observations\n",
      "1693.0 has 1 observations\n",
      "1694.0 has 3 observations\n",
      "1695.0 has 1 observations\n",
      "1699.0 has 1 observations\n",
      "1700.0 has 2 observations\n",
      "1701.0 has 3 observations\n",
      "1703.0 has 2 observations\n",
      "1704.0 has 1 observations\n",
      "1705.0 has 1 observations\n",
      "1709.0 has 2 observations\n",
      "1710.0 has 1 observations\n",
      "1711.0 has 3 observations\n",
      "1713.0 has 2 observations\n",
      "1714.0 has 1 observations\n",
      "1715.0 has 1 observations\n",
      "1719.0 has 2 observations\n",
      "1721.0 has 3 observations\n",
      "1722.0 has 1 observations\n",
      "1723.0 has 1 observations\n",
      "1729.0 has 2 observations\n",
      "1731.0 has 3 observations\n",
      "1732.0 has 1 observations\n",
      "1733.0 has 1 observations\n",
      "1739.0 has 2 observations\n",
      "1741.0 has 3 observations\n",
      "1742.0 has 1 observations\n",
      "1743.0 has 1 observations\n",
      "1748.0 has 1 observations\n",
      "1749.0 has 1 observations\n",
      "1750.0 has 2 observations\n",
      "1751.0 has 1 observations\n",
      "1752.0 has 1 observations\n",
      "1753.0 has 1 observations\n",
      "1756.0 has 19 observations\n"
     ]
    }
   ],
   "source": [
    "# check how many players' locations we have at each point in time\n",
    "# we want to make sure that after this process, all players have harmonized times and locations\n",
    "for i in data.elapsedTime.unique():\n",
    "    obs = len(data[data['elapsedTime'] == i])\n",
    "    print(f'{i} has {obs} observations')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>vehicle</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>796261</td>\n",
       "      <td>19770</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>796261</td>\n",
       "      <td>19770</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>-97954</td>\n",
       "      <td>219969</td>\n",
       "      <td>TransportAircraft</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>12.0</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>32553</td>\n",
       "      <td>273627</td>\n",
       "      <td>TransportAircraft</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>22.0</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>166787</td>\n",
       "      <td>328818</td>\n",
       "      <td>TransportAircraft</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   elapsedTime      name  location.x  location.y            vehicle\n",
       "0          0.0  AMG PUBG      796261       19770                NaN\n",
       "1          0.0  AMG PUBG      796261       19770                NaN\n",
       "2          2.0  AMG PUBG      -97954      219969  TransportAircraft\n",
       "3         12.0  AMG PUBG       32553      273627  TransportAircraft\n",
       "4         22.0  AMG PUBG      166787      328818  TransportAircraft"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we create a new dataframe with the necessary information to interpolate locations\n",
    "data_new_coords = data[['elapsedTime','name','location.x','location.y', 'vehicle']].sort_values(['name','elapsedTime']).drop_duplicates()\n",
    "data_new_coords = data_new_coords.reset_index().drop('index', axis=1)\n",
    "data_new_coords = data_new_coords.dropna(subset = ['elapsedTime'])\n",
    "# we round coordinates for ease of calculations, and because a difference in location smaller than a cm is negligible\n",
    "data_new_coords['location.x'] = data_new_coords['location.x'].apply(round)\n",
    "data_new_coords['location.y'] = data_new_coords['location.y'].apply(round)\n",
    "data_new_coords.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   0,    2,   12, ...,  386, 1704, 1714])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_new_coords.elapsedTime.unique().astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 1min 55s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# since different elapsed times don't show the location of all players, we want to interpolate the position of the players at each\n",
    "# point in time. e.g. if player a is in position (x,y) at second 10.0 and in position (x+10,y+10) at second 20.0\n",
    "# we can expect that this player will be at position (x+1,y+1) at second 11.0, and so on\n",
    "def getNewCoord(t, n):\n",
    "    aa = data_new_coords[(data_new_coords['elapsedTime'] < t) & (data_new_coords['name'] == n)] # get previous observations\n",
    "    bb = data_new_coords[(data_new_coords['elapsedTime'] > t) & (data_new_coords['name'] == n)] # get following observations\n",
    "    if (bool(aa['location.x'].any())) & (bool(bb['location.x'].any())): # if we have both previous and following observation\n",
    "        a = aa.iloc[-1] # get last previous observation\n",
    "        b = bb.iloc[0] # get first following observation\n",
    "        time = b[0] - a[0] # calculate the time difference\n",
    "        # get x and y coordinate by calculating movement speed in cm/s and get full coordinate\n",
    "        new_coord_x = a[2] + ((b[2] - a[2]) / time)*(t-a[0])  \n",
    "        new_coord_y = a[3] + ((b[3] - a[3]) / time)*(t-a[0])\n",
    "        # we also get the vehicle to check if a player has jumped from the airplane or not\n",
    "        in_aircraft = b['vehicle']\n",
    "        return new_coord_x, new_coord_y, in_aircraft\n",
    "    elif (not bool(aa['location.x'].any())) & (bool(bb['location.x'].any())):\n",
    "        # if there is no previous observation, set first following location to be current location\n",
    "        in_aircraft = b['vehicle']\n",
    "        return b.iat[0,2], b.iat[0,3], in_aircraft\n",
    "\n",
    "    # If a player doesn't have a future location, that indicates the player is dead\n",
    "    elif (bool(aa['location.x'].any())) & (not bool(bb['location.x'].any())):\n",
    "        # if there is no following observation, we assume the player has died (we hope that only in the match)\n",
    "        return 'dead', 'dead', 'dead'\n",
    "\n",
    "names=[]\n",
    "times=[]\n",
    "x_coords=[]\n",
    "y_coords=[]\n",
    "aircraft=[]\n",
    "# we will get the location every 10 seconds from second 1 (e.g. 1.0, 11.0, 21.0), and at the last second\n",
    "times_list = data_new_coords.elapsedTime.unique().astype(int)\n",
    "# we get the times in intervals of 10s from second 1 to the last second of the game\n",
    "unique_times_10s = np.arange(1, max(times_list), 10).tolist()\n",
    "# we also append the last second, regardless of the 10s interval\n",
    "unique_times_10s.append(max(times_list))\n",
    "for t in unique_times_10s:\n",
    "    for n in data_new_coords.name.unique():\n",
    "        # for every player, we get their location at these times\n",
    "        a = data_new_coords[(data_new_coords['elapsedTime'] == t) & (data_new_coords['name'] == n)]\n",
    "        # if the player has a location, we take it\n",
    "        if a['location.x'].any():\n",
    "            names.append(n)\n",
    "            times.append(t)\n",
    "            x_coords.append(a.iat[0,2])\n",
    "            y_coords.append(a.iat[0,3])\n",
    "            aircraft.append(a['vehicle'])\n",
    "        # if not, we interpolate it\n",
    "        else:\n",
    "            x_coord, y_coord, in_aircraft = getNewCoord(t,n)\n",
    "            names.append(n)\n",
    "            times.append(t)\n",
    "            x_coords.append(x_coord)\n",
    "            y_coords.append(y_coord)\n",
    "            aircraft.append(in_aircraft)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>inAircraft</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>AMG PUBG</td>\n",
       "      <td>349154</td>\n",
       "      <td>119870</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>ANX savvy</td>\n",
       "      <td>96282.7</td>\n",
       "      <td>153950</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>ASAP Drizzyy II</td>\n",
       "      <td>99480.6</td>\n",
       "      <td>152090</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>Anerge</td>\n",
       "      <td>119140</td>\n",
       "      <td>193772</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>Aviatriix</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>Azterican</td>\n",
       "      <td>237708</td>\n",
       "      <td>184558</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>BERRYWEBBLE3000</td>\n",
       "      <td>-116903</td>\n",
       "      <td>212178</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>100109</td>\n",
       "      <td>153078</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>BillsFanTim</td>\n",
       "      <td>677447</td>\n",
       "      <td>52545</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>237548</td>\n",
       "      <td>185059</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   time             name location.x location.y  inAircraft\n",
       "0     1         AMG PUBG     349154     119870           1\n",
       "1     1        ANX savvy    96282.7     153950           1\n",
       "2     1  ASAP Drizzyy II    99480.6     152090           1\n",
       "3     1           Anerge     119140     193772           1\n",
       "4     1        Aviatriix       dead       dead           0\n",
       "5     1        Azterican     237708     184558           1\n",
       "6     1  BERRYWEBBLE3000    -116903     212178           1\n",
       "7     1      BannerGnatt     100109     153078           1\n",
       "8     1      BillsFanTim     677447      52545           1\n",
       "9     1   Blky Da Ripper     237548     185059           1"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we create the dataframe from these locations\n",
    "players_location = pd.DataFrame(list(zip(times,names,x_coords, y_coords, aircraft)),\n",
    "             columns=['time','name','location.x','location.y', 'inAircraft'])\n",
    "# we also check if the players are or aren't in the aircraft with a binary variable\n",
    "players_location['inAircraft'] = players_location['inAircraft'].apply(lambda x: 1 if str(x) == 'TransportAircraft' else 0)\n",
    "players_location.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 has 100 observations\n",
      "11 has 100 observations\n",
      "21 has 100 observations\n",
      "31 has 100 observations\n",
      "41 has 100 observations\n",
      "51 has 100 observations\n",
      "61 has 100 observations\n",
      "71 has 100 observations\n",
      "81 has 100 observations\n",
      "91 has 100 observations\n",
      "101 has 100 observations\n",
      "111 has 100 observations\n",
      "121 has 100 observations\n",
      "131 has 100 observations\n",
      "141 has 100 observations\n",
      "151 has 100 observations\n",
      "161 has 100 observations\n",
      "171 has 100 observations\n",
      "181 has 100 observations\n",
      "191 has 100 observations\n",
      "201 has 100 observations\n",
      "211 has 100 observations\n",
      "221 has 100 observations\n",
      "231 has 100 observations\n",
      "241 has 100 observations\n",
      "251 has 100 observations\n",
      "261 has 100 observations\n",
      "271 has 100 observations\n",
      "281 has 100 observations\n",
      "291 has 100 observations\n",
      "301 has 100 observations\n",
      "311 has 100 observations\n",
      "321 has 100 observations\n",
      "331 has 100 observations\n",
      "341 has 100 observations\n",
      "351 has 100 observations\n",
      "361 has 100 observations\n",
      "371 has 100 observations\n",
      "381 has 100 observations\n",
      "391 has 100 observations\n",
      "401 has 100 observations\n",
      "411 has 100 observations\n",
      "421 has 100 observations\n",
      "431 has 100 observations\n",
      "441 has 100 observations\n",
      "451 has 100 observations\n",
      "461 has 100 observations\n",
      "471 has 100 observations\n",
      "481 has 100 observations\n",
      "491 has 100 observations\n",
      "501 has 100 observations\n",
      "511 has 100 observations\n",
      "521 has 100 observations\n",
      "531 has 100 observations\n",
      "541 has 100 observations\n",
      "551 has 100 observations\n",
      "561 has 100 observations\n",
      "571 has 100 observations\n",
      "581 has 100 observations\n",
      "591 has 100 observations\n",
      "601 has 100 observations\n",
      "611 has 100 observations\n",
      "621 has 100 observations\n",
      "631 has 100 observations\n",
      "641 has 100 observations\n",
      "651 has 100 observations\n",
      "661 has 100 observations\n",
      "671 has 100 observations\n",
      "681 has 100 observations\n",
      "691 has 100 observations\n",
      "701 has 100 observations\n",
      "711 has 100 observations\n",
      "721 has 100 observations\n",
      "731 has 100 observations\n",
      "741 has 100 observations\n",
      "751 has 100 observations\n",
      "761 has 100 observations\n",
      "771 has 100 observations\n",
      "781 has 100 observations\n",
      "791 has 100 observations\n",
      "801 has 100 observations\n",
      "811 has 100 observations\n",
      "821 has 100 observations\n",
      "831 has 100 observations\n",
      "841 has 100 observations\n",
      "851 has 100 observations\n",
      "861 has 100 observations\n",
      "871 has 100 observations\n",
      "881 has 100 observations\n",
      "891 has 100 observations\n",
      "901 has 100 observations\n",
      "911 has 100 observations\n",
      "921 has 100 observations\n",
      "931 has 100 observations\n",
      "941 has 100 observations\n",
      "951 has 100 observations\n",
      "961 has 100 observations\n",
      "971 has 100 observations\n",
      "981 has 100 observations\n",
      "991 has 100 observations\n",
      "1001 has 100 observations\n",
      "1011 has 100 observations\n",
      "1021 has 100 observations\n",
      "1031 has 100 observations\n",
      "1041 has 100 observations\n",
      "1051 has 100 observations\n",
      "1061 has 100 observations\n",
      "1071 has 100 observations\n",
      "1081 has 100 observations\n",
      "1091 has 100 observations\n",
      "1101 has 100 observations\n",
      "1111 has 100 observations\n",
      "1121 has 100 observations\n",
      "1131 has 100 observations\n",
      "1141 has 100 observations\n",
      "1151 has 100 observations\n",
      "1161 has 100 observations\n",
      "1171 has 100 observations\n",
      "1181 has 100 observations\n",
      "1191 has 100 observations\n",
      "1201 has 100 observations\n",
      "1211 has 100 observations\n",
      "1221 has 100 observations\n",
      "1231 has 100 observations\n",
      "1241 has 100 observations\n",
      "1251 has 100 observations\n",
      "1261 has 100 observations\n",
      "1271 has 100 observations\n",
      "1281 has 100 observations\n",
      "1291 has 100 observations\n",
      "1301 has 100 observations\n",
      "1311 has 100 observations\n",
      "1321 has 100 observations\n",
      "1331 has 100 observations\n",
      "1341 has 100 observations\n",
      "1351 has 100 observations\n",
      "1361 has 100 observations\n",
      "1371 has 100 observations\n",
      "1381 has 100 observations\n",
      "1391 has 100 observations\n",
      "1401 has 100 observations\n",
      "1411 has 100 observations\n",
      "1421 has 100 observations\n",
      "1431 has 100 observations\n",
      "1441 has 100 observations\n",
      "1451 has 100 observations\n",
      "1461 has 100 observations\n",
      "1471 has 100 observations\n",
      "1481 has 100 observations\n",
      "1491 has 100 observations\n",
      "1501 has 100 observations\n",
      "1511 has 100 observations\n",
      "1521 has 100 observations\n",
      "1531 has 100 observations\n",
      "1541 has 100 observations\n",
      "1551 has 100 observations\n",
      "1561 has 100 observations\n",
      "1571 has 100 observations\n",
      "1581 has 100 observations\n",
      "1591 has 100 observations\n",
      "1601 has 100 observations\n",
      "1611 has 100 observations\n",
      "1621 has 100 observations\n",
      "1631 has 100 observations\n",
      "1641 has 100 observations\n",
      "1651 has 100 observations\n",
      "1661 has 100 observations\n",
      "1671 has 100 observations\n",
      "1681 has 100 observations\n",
      "1691 has 100 observations\n",
      "1701 has 100 observations\n",
      "1711 has 100 observations\n",
      "1721 has 100 observations\n",
      "1731 has 100 observations\n",
      "1741 has 100 observations\n",
      "1751 has 100 observations\n",
      "1756 has 100 observations\n"
     ]
    }
   ],
   "source": [
    "# we make sure that harmonization of times and observations has been done\n",
    "for i in players_location.time.unique():\n",
    "    obs = len(players_location[players_location['time'] == i])\n",
    "    print(f'{i} has {obs} observations')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get Collaboration information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>accountId</th>\n",
       "      <th>teamId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Famous Stoner</td>\n",
       "      <td>account.23c3853ec30946e8b39a667669166f0b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>CruZifeye</td>\n",
       "      <td>account.3ec25140fdee4fd0be8343c6d90d68e6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>XxX Jav0 XxX</td>\n",
       "      <td>account.7f2470d16b544a12ade6b8d16b6cfef8</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>95</td>\n",
       "      <td>say3</td>\n",
       "      <td>account.136f6fd65c6e4bd28c9580b13a528ef8</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>96</td>\n",
       "      <td>GoldChain101</td>\n",
       "      <td>account.3b373a6839284788a506b980daed32cd</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>97</td>\n",
       "      <td>OGSKE3TBOY</td>\n",
       "      <td>account.04a9bab970a747c0bbcd99cb8fe196bf</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>98</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>account.4ada991c8de84554b6b1f32e13f5d64c</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>99</td>\n",
       "      <td>ASAP Drizzyy II</td>\n",
       "      <td>account.94a181525175440b99896ef6494e1ee6</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               name                                 accountId  teamId\n",
       "0    Blky Da Ripper  account.3e6ca222bda44f21bdb811a32c3aa21f       1\n",
       "1     Famous Stoner  account.23c3853ec30946e8b39a667669166f0b       1\n",
       "2         CruZifeye  account.3ec25140fdee4fd0be8343c6d90d68e6       1\n",
       "3          walas-ii  account.f915e33e51374cb7888f485a91331417       1\n",
       "4      XxX Jav0 XxX  account.7f2470d16b544a12ade6b8d16b6cfef8       2\n",
       "..              ...                                       ...     ...\n",
       "95             say3  account.136f6fd65c6e4bd28c9580b13a528ef8      26\n",
       "96     GoldChain101  account.3b373a6839284788a506b980daed32cd      26\n",
       "97       OGSKE3TBOY  account.04a9bab970a747c0bbcd99cb8fe196bf      27\n",
       "98      BannerGnatt  account.4ada991c8de84554b6b1f32e13f5d64c      27\n",
       "99  ASAP Drizzyy II  account.94a181525175440b99896ef6494e1ee6      27\n",
       "\n",
       "[100 rows x 3 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# encode player names\n",
    "# we do this to get the teamIds in the main dataframe\n",
    "players = data[['name', 'accountId','teamId']].drop_duplicates()\n",
    "players = players.sort_values('teamId').reset_index().drop(['index'], axis=1)\n",
    "players"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>inAircraft</th>\n",
       "      <th>accountId</th>\n",
       "      <th>teamId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>237548</td>\n",
       "      <td>185059</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>CruZifeye</td>\n",
       "      <td>198981</td>\n",
       "      <td>187710</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3ec25140fdee4fd0be8343c6d90d68e6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Famous Stoner</td>\n",
       "      <td>237436</td>\n",
       "      <td>184385</td>\n",
       "      <td>1</td>\n",
       "      <td>account.23c3853ec30946e8b39a667669166f0b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>-123677</td>\n",
       "      <td>209394</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>12695.5</td>\n",
       "      <td>265463</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   time            name location.x location.y  inAircraft  \\\n",
       "0     1  Blky Da Ripper     237548     185059           1   \n",
       "1     1       CruZifeye     198981     187710           1   \n",
       "2     1   Famous Stoner     237436     184385           1   \n",
       "3     1        walas-ii    -123677     209394           0   \n",
       "4    11  Blky Da Ripper    12695.5     265463           1   \n",
       "\n",
       "                                  accountId  teamId  \n",
       "0  account.3e6ca222bda44f21bdb811a32c3aa21f       1  \n",
       "1  account.3ec25140fdee4fd0be8343c6d90d68e6       1  \n",
       "2  account.23c3853ec30946e8b39a667669166f0b       1  \n",
       "3  account.f915e33e51374cb7888f485a91331417       1  \n",
       "4  account.3e6ca222bda44f21bdb811a32c3aa21f       1  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we merge these names to get accountId and teamId in the dataframe\n",
    "players_location = pd.merge(players_location,\n",
    "                            players,\n",
    "                            left_on = 'name',\n",
    "                            right_on = 'name',\n",
    "                            how = 'left')\n",
    "# moreover, since we are goint to later \"asign\" a number to the players (e.g. player1, player2), we want to make sure\n",
    "# that the criteria is consistent for all teams. Therefore, we do it with the .sort_values method\n",
    "# players on each team are ordered with this method inside the team\n",
    "players_location = players_location[players_location['time'] >= 0].sort_values(['teamId','time', 'name']).reset_index().drop(['index'], axis=1)\n",
    "players_location.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting the Centroids and distance to centroid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we want to get the centroids of locations of the players in a team for every elapsed time\n",
    "# for this, we create this function\n",
    "def getCentroid(*args):\n",
    "    \"Returns a single coordinate centroid\"\n",
    "    # the function gets all the 1D locations of the players and computes the mean\n",
    "    # we use *args so that we don't need to specify the number of locations\n",
    "    # moreover, we only take the players whose location is not 'dead'\n",
    "    locs = [i for i in [*args[0]] if type(i) != str]\n",
    "    s = sum(locs)\n",
    "    n = len(locs)\n",
    "    # if there is no player alive, the location returned will be 0\n",
    "    if n == 0:\n",
    "        return 0\n",
    "    else:\n",
    "        return s/n\n",
    "teams = []\n",
    "times = []\n",
    "centroidsx = []\n",
    "centroidsy = []\n",
    "# now, we apply this to all team/time pairs\n",
    "for teamId in players_location.teamId.unique():\n",
    "    for time in players_location.time.unique():\n",
    "        df = players_location[(players_location.teamId == teamId) & (players_location.time == time)]\n",
    "        centroidx = getCentroid(df['location.x'])\n",
    "        centroidy = getCentroid(df['location.y'])\n",
    "        teams.append(teamId)\n",
    "        times.append(time)\n",
    "        centroidsx.append(centroidx)\n",
    "        centroidsy.append(centroidy)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we turn this into a dataframe\n",
    "dict = {'teamId':teams, 'time':times, 'centroid.x':centroidsx, 'centroid.y':centroidsy}\n",
    "players_location = pd.merge(players_location, pd.DataFrame(dict), how = 'left', on = ['teamId','time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>inAircraft</th>\n",
       "      <th>accountId</th>\n",
       "      <th>teamId</th>\n",
       "      <th>centroid.x</th>\n",
       "      <th>centroid.y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>80</td>\n",
       "      <td>201</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>435685</td>\n",
       "      <td>338991</td>\n",
       "      <td>0</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "      <td>437500.266667</td>\n",
       "      <td>336476.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>81</td>\n",
       "      <td>201</td>\n",
       "      <td>CruZifeye</td>\n",
       "      <td>440985</td>\n",
       "      <td>331676</td>\n",
       "      <td>0</td>\n",
       "      <td>account.3ec25140fdee4fd0be8343c6d90d68e6</td>\n",
       "      <td>1</td>\n",
       "      <td>437500.266667</td>\n",
       "      <td>336476.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>82</td>\n",
       "      <td>201</td>\n",
       "      <td>Famous Stoner</td>\n",
       "      <td>435831</td>\n",
       "      <td>338763</td>\n",
       "      <td>0</td>\n",
       "      <td>account.23c3853ec30946e8b39a667669166f0b</td>\n",
       "      <td>1</td>\n",
       "      <td>437500.266667</td>\n",
       "      <td>336476.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>83</td>\n",
       "      <td>201</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>1</td>\n",
       "      <td>437500.266667</td>\n",
       "      <td>336476.666667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    time            name location.x location.y  inAircraft  \\\n",
       "80   201  Blky Da Ripper     435685     338991           0   \n",
       "81   201       CruZifeye     440985     331676           0   \n",
       "82   201   Famous Stoner     435831     338763           0   \n",
       "83   201        walas-ii       dead       dead           0   \n",
       "\n",
       "                                   accountId  teamId     centroid.x  \\\n",
       "80  account.3e6ca222bda44f21bdb811a32c3aa21f       1  437500.266667   \n",
       "81  account.3ec25140fdee4fd0be8343c6d90d68e6       1  437500.266667   \n",
       "82  account.23c3853ec30946e8b39a667669166f0b       1  437500.266667   \n",
       "83  account.f915e33e51374cb7888f485a91331417       1  437500.266667   \n",
       "\n",
       "       centroid.y  \n",
       "80  336476.666667  \n",
       "81  336476.666667  \n",
       "82  336476.666667  \n",
       "83  336476.666667  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# and we merge it with the original dataframe\n",
    "players_location[(players_location.teamId == 1) & (players_location.time == 201)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# now we want to calculate the distance of the players to this centroid\n",
    "def getDistance(locx_1, locy_1, locx_2, locy_2):\n",
    "    \"\"\"\n",
    "    Returns the euclidean distance between two points in a plane given their two coordinates, x and y\n",
    "    \"\"\"\n",
    "    try:\n",
    "        return ((locx_2 - locx_1)**2 + (locy_2 - locy_1)**2)**(1/2)\n",
    "    # if player is dead ('dead' in the location), we return a negative distance\n",
    "    except:\n",
    "        return -999"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we apply the getDistance function to all rows\n",
    "players_location['distance_centroid'] = np.vectorize(getDistance)(players_location['location.x'], players_location['location.y'], players_location['centroid.x'], players_location['centroid.y'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting individual distances between players"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get distance between players in the same team for all times\n",
    "# first, we create an empty numpy array with 4 columns (1 per player)\n",
    "distances = np.empty((0,4))\n",
    "for team in players_location.teamId.unique():\n",
    "    # we select the data corresponding to the team\n",
    "    df1 = players_location[players_location.teamId == team]\n",
    "    players = {}\n",
    "    # I have to check why I did this in the first place because I think it might not be useful anymore\n",
    "    for i, player in zip(range(len(data.name.unique())), data.name.unique()):\n",
    "        players[f'player{i+1}'] = player\n",
    "    for time in df1.time.unique():\n",
    "        # and we select the data of the corresponding time\n",
    "        df2 = df1[df1['time'] == time]\n",
    "        # if there are more than 4 players, it's because of a bug where a player disconnects before the match starts\n",
    "        # so another player gets in, and the data shows 5 players, but 1 shows no location (\"dead\")\n",
    "        # we want to drop this player\n",
    "        # for that, we store their name and drop it before every computation\n",
    "        if (len(df2) > 4) and (time==1):\n",
    "            name = df2[df2.applymap(type)['location.x'] == str]['name'].iloc[0]\n",
    "        if len(df2) > 4:\n",
    "            df2 = df2[df2['name'] != name]\n",
    "        # get a diagonal matrix with the distances between every pair of players from the same team in a time\n",
    "        # including themselves\n",
    "        d = np.array([[getDistance(i,j,k,l) for i,j in zip(df2['location.x'], df2['location.y'])] \n",
    "                      for k,l in zip(df2['location.x'], df2['location.y'])])\n",
    "        if d.shape[1] == 4:\n",
    "            distances = np.concatenate((distances, d))\n",
    "        # if the team has less than 4 players, we add a Null value for the column of the non-existent player\n",
    "        # this is just for the purpose of maintaining the shape of the matrix\n",
    "        else:\n",
    "            h = np.full((d.shape[1], 4 - d.shape[1]), np.nan)\n",
    "            d = np.column_stack((d, h))\n",
    "            distances = np.concatenate((distances, d))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# then, we merge it with the main dataframe\n",
    "players_location = pd.concat((players_location, pd.DataFrame(distances, columns = ['player1','player2','player3','player4'])),\n",
    "                            axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "players_location.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting Clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import DBSCAN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 406,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2667</td>\n",
       "      <td>207235</td>\n",
       "      <td>376063</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2668</td>\n",
       "      <td>416834</td>\n",
       "      <td>431625</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2669</td>\n",
       "      <td>198193</td>\n",
       "      <td>361346</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     location.x location.y  cluster\n",
       "2667     207235     376063        0\n",
       "2668     416834     431625        1\n",
       "2669     198193     361346        2"
      ]
     },
     "execution_count": 406,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = players_location[(players_location.teamId == 5) & (players_location.time == 41)]\n",
    "x = players_location[(players_location.teamId == 5) & (players_location.time == 41)][['location.x', 'location.y']]\n",
    "clustering = DBSCAN(eps = 10000, min_samples = 1.0).fit(x)\n",
    "x['cluster'] = clustering.labels_\n",
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 39.8 s\n"
     ]
    }
   ],
   "source": [
    "# now we want to create clusters of players in the team at every point in time\n",
    "%%time\n",
    "times = []\n",
    "names = []\n",
    "clusters = []\n",
    "for teamId in players_location.teamId.unique():\n",
    "    for time in players_location.time.unique():\n",
    "        # given a team and a time, we get their locations and apply the DBSCAN algorithm\n",
    "        df = players_location[(players_location.teamId == teamId) & (players_location.time == time)]\n",
    "        X = df[['location.x','location.y']]\n",
    "        X = X.applymap(lambda x: -999 if x == 'dead' else int(x))\n",
    "        # the algorithm is applied with a minimum of 1 player in the cluster, and a maximum distance of 100m between players\n",
    "        clustering = DBSCAN(eps=10000, min_samples=1).fit(X)\n",
    "        times = np.concatenate((times, df['time'].tolist())).tolist()\n",
    "        names = np.concatenate((names, df['name'].tolist())).tolist()\n",
    "        clusters = np.concatenate((clusters, clustering.labels_.tolist())).tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>inAircraft</th>\n",
       "      <th>accountId</th>\n",
       "      <th>teamId</th>\n",
       "      <th>centroid.x</th>\n",
       "      <th>centroid.y</th>\n",
       "      <th>distance_centroid</th>\n",
       "      <th>player1</th>\n",
       "      <th>player2</th>\n",
       "      <th>player3</th>\n",
       "      <th>player4</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>237548</td>\n",
       "      <td>185059</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>100191.660160</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>38657.482536</td>\n",
       "      <td>683.325737</td>\n",
       "      <td>362043.273139</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>CruZifeye</td>\n",
       "      <td>198981</td>\n",
       "      <td>187710</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3ec25140fdee4fd0be8343c6d90d68e6</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>61534.446310</td>\n",
       "      <td>38657.482536</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>38598.970855</td>\n",
       "      <td>323385.830358</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Famous Stoner</td>\n",
       "      <td>237436</td>\n",
       "      <td>184385</td>\n",
       "      <td>1</td>\n",
       "      <td>account.23c3853ec30946e8b39a667669166f0b</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>100127.475625</td>\n",
       "      <td>683.325737</td>\n",
       "      <td>38598.970855</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>361978.483432</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>-123677</td>\n",
       "      <td>209394</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>261851.782941</td>\n",
       "      <td>362043.273139</td>\n",
       "      <td>323385.830358</td>\n",
       "      <td>361978.483432</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>12695.5</td>\n",
       "      <td>265463</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "      <td>13262.55</td>\n",
       "      <td>265696.150000</td>\n",
       "      <td>613.186697</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2066.128004</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>386.618792</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17695</td>\n",
       "      <td>1751</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.4ada991c8de84554b6b1f32e13f5d64c</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17696</td>\n",
       "      <td>1751</td>\n",
       "      <td>OGSKE3TBOY</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.04a9bab970a747c0bbcd99cb8fe196bf</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17697</td>\n",
       "      <td>1756</td>\n",
       "      <td>ASAP Drizzyy II</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.94a181525175440b99896ef6494e1ee6</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17698</td>\n",
       "      <td>1756</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.4ada991c8de84554b6b1f32e13f5d64c</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17699</td>\n",
       "      <td>1756</td>\n",
       "      <td>OGSKE3TBOY</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.04a9bab970a747c0bbcd99cb8fe196bf</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>17700 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       time             name location.x location.y  inAircraft  \\\n",
       "0         1   Blky Da Ripper     237548     185059           1   \n",
       "1         1        CruZifeye     198981     187710           1   \n",
       "2         1    Famous Stoner     237436     184385           1   \n",
       "3         1         walas-ii    -123677     209394           0   \n",
       "4        11   Blky Da Ripper    12695.5     265463           1   \n",
       "...     ...              ...        ...        ...         ...   \n",
       "17695  1751      BannerGnatt       dead       dead           0   \n",
       "17696  1751       OGSKE3TBOY       dead       dead           0   \n",
       "17697  1756  ASAP Drizzyy II       dead       dead           0   \n",
       "17698  1756      BannerGnatt       dead       dead           0   \n",
       "17699  1756       OGSKE3TBOY       dead       dead           0   \n",
       "\n",
       "                                      accountId  teamId  centroid.x  \\\n",
       "0      account.3e6ca222bda44f21bdb811a32c3aa21f       1   137572.00   \n",
       "1      account.3ec25140fdee4fd0be8343c6d90d68e6       1   137572.00   \n",
       "2      account.23c3853ec30946e8b39a667669166f0b       1   137572.00   \n",
       "3      account.f915e33e51374cb7888f485a91331417       1   137572.00   \n",
       "4      account.3e6ca222bda44f21bdb811a32c3aa21f       1    13262.55   \n",
       "...                                         ...     ...         ...   \n",
       "17695  account.4ada991c8de84554b6b1f32e13f5d64c      27        0.00   \n",
       "17696  account.04a9bab970a747c0bbcd99cb8fe196bf      27        0.00   \n",
       "17697  account.94a181525175440b99896ef6494e1ee6      27        0.00   \n",
       "17698  account.4ada991c8de84554b6b1f32e13f5d64c      27        0.00   \n",
       "17699  account.04a9bab970a747c0bbcd99cb8fe196bf      27        0.00   \n",
       "\n",
       "          centroid.y  distance_centroid        player1        player2  \\\n",
       "0      191636.854167      100191.660160       0.000000   38657.482536   \n",
       "1      191636.854167       61534.446310   38657.482536       0.000000   \n",
       "2      191636.854167      100127.475625     683.325737   38598.970855   \n",
       "3      191636.854167      261851.782941  362043.273139  323385.830358   \n",
       "4      265696.150000         613.186697       0.000000    2066.128004   \n",
       "...              ...                ...            ...            ...   \n",
       "17695       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17696       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17697       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17698       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17699       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "\n",
       "             player3        player4  cluster  \n",
       "0         683.325737  362043.273139      0.0  \n",
       "1       38598.970855  323385.830358      1.0  \n",
       "2           0.000000  361978.483432      0.0  \n",
       "3      361978.483432       0.000000      2.0  \n",
       "4           0.000000     386.618792      0.0  \n",
       "...              ...            ...      ...  \n",
       "17695    -999.000000            NaN     -1.0  \n",
       "17696    -999.000000            NaN     -1.0  \n",
       "17697    -999.000000            NaN     -1.0  \n",
       "17698    -999.000000            NaN     -1.0  \n",
       "17699    -999.000000            NaN     -1.0  \n",
       "\n",
       "[17700 rows x 15 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# then, we merge it into the main dataframe\n",
    "players_location = pd.merge(players_location, pd.DataFrame(list(zip(times,names,clusters)), columns = ['time','name','cluster']),\n",
    "                            how = 'left', on = ['time','name'])\n",
    "# for the dead players, we create a separate cluster, since we will drop this cluster later\n",
    "players_location.loc[players_location['location.x'] == 'dead', 'cluster'] = -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>inAircraft</th>\n",
       "      <th>accountId</th>\n",
       "      <th>teamId</th>\n",
       "      <th>centroid.x</th>\n",
       "      <th>centroid.y</th>\n",
       "      <th>distance_centroid</th>\n",
       "      <th>player1</th>\n",
       "      <th>player2</th>\n",
       "      <th>player3</th>\n",
       "      <th>player4</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>237548</td>\n",
       "      <td>185059</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>100191.660160</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>38657.482536</td>\n",
       "      <td>683.325737</td>\n",
       "      <td>362043.273139</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>CruZifeye</td>\n",
       "      <td>198981</td>\n",
       "      <td>187710</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3ec25140fdee4fd0be8343c6d90d68e6</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>61534.446310</td>\n",
       "      <td>38657.482536</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>38598.970855</td>\n",
       "      <td>323385.830358</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Famous Stoner</td>\n",
       "      <td>237436</td>\n",
       "      <td>184385</td>\n",
       "      <td>1</td>\n",
       "      <td>account.23c3853ec30946e8b39a667669166f0b</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>100127.475625</td>\n",
       "      <td>683.325737</td>\n",
       "      <td>38598.970855</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>361978.483432</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>walas-ii</td>\n",
       "      <td>-123677</td>\n",
       "      <td>209394</td>\n",
       "      <td>0</td>\n",
       "      <td>account.f915e33e51374cb7888f485a91331417</td>\n",
       "      <td>1</td>\n",
       "      <td>137572.00</td>\n",
       "      <td>191636.854167</td>\n",
       "      <td>261851.782941</td>\n",
       "      <td>362043.273139</td>\n",
       "      <td>323385.830358</td>\n",
       "      <td>361978.483432</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>11</td>\n",
       "      <td>Blky Da Ripper</td>\n",
       "      <td>12695.5</td>\n",
       "      <td>265463</td>\n",
       "      <td>1</td>\n",
       "      <td>account.3e6ca222bda44f21bdb811a32c3aa21f</td>\n",
       "      <td>1</td>\n",
       "      <td>13262.55</td>\n",
       "      <td>265696.150000</td>\n",
       "      <td>613.186697</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2066.128004</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>386.618792</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17695</td>\n",
       "      <td>1751</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.4ada991c8de84554b6b1f32e13f5d64c</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17696</td>\n",
       "      <td>1751</td>\n",
       "      <td>OGSKE3TBOY</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.04a9bab970a747c0bbcd99cb8fe196bf</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17697</td>\n",
       "      <td>1756</td>\n",
       "      <td>ASAP Drizzyy II</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.94a181525175440b99896ef6494e1ee6</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17698</td>\n",
       "      <td>1756</td>\n",
       "      <td>BannerGnatt</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.4ada991c8de84554b6b1f32e13f5d64c</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17699</td>\n",
       "      <td>1756</td>\n",
       "      <td>OGSKE3TBOY</td>\n",
       "      <td>dead</td>\n",
       "      <td>dead</td>\n",
       "      <td>0</td>\n",
       "      <td>account.04a9bab970a747c0bbcd99cb8fe196bf</td>\n",
       "      <td>27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Z</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>17700 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       time             name location.x location.y  inAircraft  \\\n",
       "0         1   Blky Da Ripper     237548     185059           1   \n",
       "1         1        CruZifeye     198981     187710           1   \n",
       "2         1    Famous Stoner     237436     184385           1   \n",
       "3         1         walas-ii    -123677     209394           0   \n",
       "4        11   Blky Da Ripper    12695.5     265463           1   \n",
       "...     ...              ...        ...        ...         ...   \n",
       "17695  1751      BannerGnatt       dead       dead           0   \n",
       "17696  1751       OGSKE3TBOY       dead       dead           0   \n",
       "17697  1756  ASAP Drizzyy II       dead       dead           0   \n",
       "17698  1756      BannerGnatt       dead       dead           0   \n",
       "17699  1756       OGSKE3TBOY       dead       dead           0   \n",
       "\n",
       "                                      accountId  teamId  centroid.x  \\\n",
       "0      account.3e6ca222bda44f21bdb811a32c3aa21f       1   137572.00   \n",
       "1      account.3ec25140fdee4fd0be8343c6d90d68e6       1   137572.00   \n",
       "2      account.23c3853ec30946e8b39a667669166f0b       1   137572.00   \n",
       "3      account.f915e33e51374cb7888f485a91331417       1   137572.00   \n",
       "4      account.3e6ca222bda44f21bdb811a32c3aa21f       1    13262.55   \n",
       "...                                         ...     ...         ...   \n",
       "17695  account.4ada991c8de84554b6b1f32e13f5d64c      27        0.00   \n",
       "17696  account.04a9bab970a747c0bbcd99cb8fe196bf      27        0.00   \n",
       "17697  account.94a181525175440b99896ef6494e1ee6      27        0.00   \n",
       "17698  account.4ada991c8de84554b6b1f32e13f5d64c      27        0.00   \n",
       "17699  account.04a9bab970a747c0bbcd99cb8fe196bf      27        0.00   \n",
       "\n",
       "          centroid.y  distance_centroid        player1        player2  \\\n",
       "0      191636.854167      100191.660160       0.000000   38657.482536   \n",
       "1      191636.854167       61534.446310   38657.482536       0.000000   \n",
       "2      191636.854167      100127.475625     683.325737   38598.970855   \n",
       "3      191636.854167      261851.782941  362043.273139  323385.830358   \n",
       "4      265696.150000         613.186697       0.000000    2066.128004   \n",
       "...              ...                ...            ...            ...   \n",
       "17695       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17696       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17697       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17698       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "17699       0.000000        -999.000000    -999.000000    -999.000000   \n",
       "\n",
       "             player3        player4 cluster  \n",
       "0         683.325737  362043.273139       A  \n",
       "1       38598.970855  323385.830358       B  \n",
       "2           0.000000  361978.483432       A  \n",
       "3      361978.483432       0.000000       C  \n",
       "4           0.000000     386.618792       A  \n",
       "...              ...            ...     ...  \n",
       "17695    -999.000000            NaN       Z  \n",
       "17696    -999.000000            NaN       Z  \n",
       "17697    -999.000000            NaN       Z  \n",
       "17698    -999.000000            NaN       Z  \n",
       "17699    -999.000000            NaN       Z  \n",
       "\n",
       "[17700 rows x 15 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we map the clusters to a lable\n",
    "l_clusters = {0.0:'A', 1.0:'B', 2.0:'C', 3.0:'D', -1.0:'Z'}\n",
    "players_location.cluster = players_location.cluster.map(l_clusters)\n",
    "players_location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>cluster</th>\n",
       "      <th>time</th>\n",
       "      <th>teamId</th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>Z</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4774</td>\n",
       "      <td>1756</td>\n",
       "      <td>23</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4775</td>\n",
       "      <td>1756</td>\n",
       "      <td>24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4776</td>\n",
       "      <td>1756</td>\n",
       "      <td>25</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4777</td>\n",
       "      <td>1756</td>\n",
       "      <td>26</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4778</td>\n",
       "      <td>1756</td>\n",
       "      <td>27</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4779 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "cluster  time  teamId    A    B    C    D    Z\n",
       "0           1       1  2.0  1.0  1.0  0.0  0.0\n",
       "1           1       2  3.0  0.0  0.0  0.0  0.0\n",
       "2           1       3  2.0  1.0  1.0  0.0  0.0\n",
       "3           1       4  3.0  1.0  0.0  0.0  0.0\n",
       "4           1       5  1.0  2.0  0.0  0.0  0.0\n",
       "...       ...     ...  ...  ...  ...  ...  ...\n",
       "4774     1756      23  0.0  0.0  0.0  0.0  3.0\n",
       "4775     1756      24  0.0  0.0  0.0  0.0  4.0\n",
       "4776     1756      25  0.0  0.0  0.0  0.0  3.0\n",
       "4777     1756      26  0.0  0.0  0.0  0.0  4.0\n",
       "4778     1756      27  0.0  0.0  0.0  0.0  3.0\n",
       "\n",
       "[4779 rows x 7 columns]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# now, we want to count the number of players inside each cluster\n",
    "# we will store this to merge it with the final data later\n",
    "cluster_count = players_location.pivot_table(values = ['name'], index = ['time', 'teamId'], columns=['cluster'], aggfunc = 'count')\n",
    "cluster_count.columns = cluster_count.columns.droplevel()\n",
    "cluster_count = cluster_count.reset_index()\n",
    "cluster_count = cluster_count.fillna(0)\n",
    "cluster_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Flatten Dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get the distances in flat format for each team\n",
    "rows = []\n",
    "for teamId in players_location.teamId.unique():\n",
    "    n_players = len(players_location[players_location['teamId'] == teamId].name.unique())\n",
    "    for time in players_location.time.unique():\n",
    "        df = players_location[(players_location['teamId'] == teamId) & (players_location['time'] == time)]\n",
    "        in_aircraft = sum(df['inAircraft'])\n",
    "        if df.shape[0] == 4:\n",
    "            row = (time, teamId, n_players, in_aircraft, \n",
    "                   df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], df.iat[2,13],\n",
    "                   df.iat[0,9], df.iat[1,9], df.iat[2,9], df.iat[3,9])\n",
    "        # teams with less than 4 players need to be inputed some nans in some columns\n",
    "        elif df.shape[0] == 3:\n",
    "            row = (time, teamId, n_players, in_aircraft, \n",
    "                   df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], df.iat[2,13],\n",
    "                   df.iat[0,9], df.iat[1,9], df.iat[2,9], np.nan)            \n",
    "        elif df.shape[0] == 2:\n",
    "            row = (time, teamId, n_players, in_aircraft, \n",
    "                   df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], np.nan,\n",
    "                   df.iat[0,9], df.iat[1,9], np.nan, np.nan)\n",
    "        else:\n",
    "            row = (time, teamId, n_players, in_aircraft, \n",
    "                   df.iat[0,11], df.iat[0,12], df.iat[0,13], np.nan, np.nan, np.nan,\n",
    "                   df.iat[0,9], np.nan, np.nan, np.nan)\n",
    "        rows.append(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>teamId</th>\n",
       "      <th>n_players</th>\n",
       "      <th>in_aircraft</th>\n",
       "      <th>distance1_2</th>\n",
       "      <th>distance1_3</th>\n",
       "      <th>distance1_4</th>\n",
       "      <th>distance2_3</th>\n",
       "      <th>distance2_4</th>\n",
       "      <th>distance3_4</th>\n",
       "      <th>distance_centroid1</th>\n",
       "      <th>distance_centroid2</th>\n",
       "      <th>distance_centroid3</th>\n",
       "      <th>distance_centroid4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>38657.482536</td>\n",
       "      <td>683.325737</td>\n",
       "      <td>362043.273139</td>\n",
       "      <td>38598.970855</td>\n",
       "      <td>323385.830358</td>\n",
       "      <td>361978.483432</td>\n",
       "      <td>100191.660160</td>\n",
       "      <td>61534.446310</td>\n",
       "      <td>100127.475625</td>\n",
       "      <td>261851.782941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2066.128004</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>386.618792</td>\n",
       "      <td>2066.128004</td>\n",
       "      <td>1679.509226</td>\n",
       "      <td>386.618792</td>\n",
       "      <td>613.186697</td>\n",
       "      <td>1452.941308</td>\n",
       "      <td>613.186697</td>\n",
       "      <td>226.567926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1774.566364</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>165.180689</td>\n",
       "      <td>1774.566364</td>\n",
       "      <td>1609.386032</td>\n",
       "      <td>165.180689</td>\n",
       "      <td>484.936689</td>\n",
       "      <td>1289.629684</td>\n",
       "      <td>484.936689</td>\n",
       "      <td>319.756411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>31</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>6989.823581</td>\n",
       "      <td>4978.102058</td>\n",
       "      <td>336.218396</td>\n",
       "      <td>2012.300417</td>\n",
       "      <td>7326.041400</td>\n",
       "      <td>5314.304460</td>\n",
       "      <td>2907.903694</td>\n",
       "      <td>4081.951379</td>\n",
       "      <td>2070.273167</td>\n",
       "      <td>3244.117611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>113952.915426</td>\n",
       "      <td>101169.937339</td>\n",
       "      <td>101030.109283</td>\n",
       "      <td>13069.506146</td>\n",
       "      <td>13127.803355</td>\n",
       "      <td>409.874086</td>\n",
       "      <td>79033.141467</td>\n",
       "      <td>34948.754662</td>\n",
       "      <td>22156.149667</td>\n",
       "      <td>22004.737349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4774</td>\n",
       "      <td>1721</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4775</td>\n",
       "      <td>1731</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4776</td>\n",
       "      <td>1741</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4777</td>\n",
       "      <td>1751</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4778</td>\n",
       "      <td>1756</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4779 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      time  teamId  n_players  in_aircraft    distance1_2    distance1_3  \\\n",
       "0        1       1          4            3   38657.482536     683.325737   \n",
       "1       11       1          4            3    2066.128004       0.000000   \n",
       "2       21       1          4            3    1774.566364       0.000000   \n",
       "3       31       1          4            1    6989.823581    4978.102058   \n",
       "4       41       1          4            1  113952.915426  101169.937339   \n",
       "...    ...     ...        ...          ...            ...            ...   \n",
       "4774  1721      27          3            0    -999.000000    -999.000000   \n",
       "4775  1731      27          3            0    -999.000000    -999.000000   \n",
       "4776  1741      27          3            0    -999.000000    -999.000000   \n",
       "4777  1751      27          3            0    -999.000000    -999.000000   \n",
       "4778  1756      27          3            0    -999.000000    -999.000000   \n",
       "\n",
       "        distance1_4   distance2_3    distance2_4    distance3_4  \\\n",
       "0     362043.273139  38598.970855  323385.830358  361978.483432   \n",
       "1        386.618792   2066.128004    1679.509226     386.618792   \n",
       "2        165.180689   1774.566364    1609.386032     165.180689   \n",
       "3        336.218396   2012.300417    7326.041400    5314.304460   \n",
       "4     101030.109283  13069.506146   13127.803355     409.874086   \n",
       "...             ...           ...            ...            ...   \n",
       "4774            NaN   -999.000000            NaN            NaN   \n",
       "4775            NaN   -999.000000            NaN            NaN   \n",
       "4776            NaN   -999.000000            NaN            NaN   \n",
       "4777            NaN   -999.000000            NaN            NaN   \n",
       "4778            NaN   -999.000000            NaN            NaN   \n",
       "\n",
       "      distance_centroid1  distance_centroid2  distance_centroid3  \\\n",
       "0          100191.660160        61534.446310       100127.475625   \n",
       "1             613.186697         1452.941308          613.186697   \n",
       "2             484.936689         1289.629684          484.936689   \n",
       "3            2907.903694         4081.951379         2070.273167   \n",
       "4           79033.141467        34948.754662        22156.149667   \n",
       "...                  ...                 ...                 ...   \n",
       "4774         -999.000000         -999.000000         -999.000000   \n",
       "4775         -999.000000         -999.000000         -999.000000   \n",
       "4776         -999.000000         -999.000000         -999.000000   \n",
       "4777         -999.000000         -999.000000         -999.000000   \n",
       "4778         -999.000000         -999.000000         -999.000000   \n",
       "\n",
       "      distance_centroid4  \n",
       "0          261851.782941  \n",
       "1             226.567926  \n",
       "2             319.756411  \n",
       "3            3244.117611  \n",
       "4           22004.737349  \n",
       "...                  ...  \n",
       "4774                 NaN  \n",
       "4775                 NaN  \n",
       "4776                 NaN  \n",
       "4777                 NaN  \n",
       "4778                 NaN  \n",
       "\n",
       "[4779 rows x 14 columns]"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# then, we create the final dataframe with the flattened columns\n",
    "columns = ['time', 'teamId', 'n_players', 'in_aircraft', \n",
    "           'distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4',\n",
    "           'distance_centroid1', 'distance_centroid2', 'distance_centroid3', 'distance_centroid4']\n",
    "team_distances = pd.DataFrame(rows, columns=columns)\n",
    "team_distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "# finally, we merge it with the clusters dataframe and we are done!\n",
    "team_distances = pd.merge(team_distances, cluster_count, how='left', on=['time','teamId']).drop('Z', axis=1)\n",
    "team_distances = team_distances.rename(columns = {'A': 'cluster_A', 'B': 'cluster_B', 'C': 'cluster_C', 'D': 'cluster_D'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>teamId</th>\n",
       "      <th>n_players</th>\n",
       "      <th>in_aircraft</th>\n",
       "      <th>distance1_2</th>\n",
       "      <th>distance1_3</th>\n",
       "      <th>distance1_4</th>\n",
       "      <th>distance2_3</th>\n",
       "      <th>distance2_4</th>\n",
       "      <th>distance3_4</th>\n",
       "      <th>distance_centroid1</th>\n",
       "      <th>distance_centroid2</th>\n",
       "      <th>distance_centroid3</th>\n",
       "      <th>distance_centroid4</th>\n",
       "      <th>cluster_A</th>\n",
       "      <th>cluster_B</th>\n",
       "      <th>cluster_C</th>\n",
       "      <th>cluster_D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>4602</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1171.241842</td>\n",
       "      <td>540.915500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1063.893586</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>494.067528</td>\n",
       "      <td>723.783612</td>\n",
       "      <td>405.121495</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4603</td>\n",
       "      <td>11</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>549.779149</td>\n",
       "      <td>1102.339661</td>\n",
       "      <td>NaN</td>\n",
       "      <td>552.560594</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>550.706261</td>\n",
       "      <td>0.937820</td>\n",
       "      <td>551.633409</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4604</td>\n",
       "      <td>21</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>11064.808583</td>\n",
       "      <td>6727.643228</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4337.165372</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5930.817269</td>\n",
       "      <td>5133.991316</td>\n",
       "      <td>796.825977</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4605</td>\n",
       "      <td>31</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>8173.311523</td>\n",
       "      <td>44066.275023</td>\n",
       "      <td>NaN</td>\n",
       "      <td>36001.700680</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17388.236060</td>\n",
       "      <td>9333.418760</td>\n",
       "      <td>26685.649642</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4606</td>\n",
       "      <td>41</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>20784.781912</td>\n",
       "      <td>106103.333871</td>\n",
       "      <td>NaN</td>\n",
       "      <td>85787.088425</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>42190.605451</td>\n",
       "      <td>21921.440394</td>\n",
       "      <td>63946.745933</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4774</td>\n",
       "      <td>1721</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4775</td>\n",
       "      <td>1731</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4776</td>\n",
       "      <td>1741</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4777</td>\n",
       "      <td>1751</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4778</td>\n",
       "      <td>1756</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>177 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      time  teamId  n_players  in_aircraft   distance1_2    distance1_3  \\\n",
       "4602     1      27          3            3   1171.241842     540.915500   \n",
       "4603    11      27          3            3    549.779149    1102.339661   \n",
       "4604    21      27          3            3  11064.808583    6727.643228   \n",
       "4605    31      27          3            1   8173.311523   44066.275023   \n",
       "4606    41      27          3            0  20784.781912  106103.333871   \n",
       "...    ...     ...        ...          ...           ...            ...   \n",
       "4774  1721      27          3            0   -999.000000    -999.000000   \n",
       "4775  1731      27          3            0   -999.000000    -999.000000   \n",
       "4776  1741      27          3            0   -999.000000    -999.000000   \n",
       "4777  1751      27          3            0   -999.000000    -999.000000   \n",
       "4778  1756      27          3            0   -999.000000    -999.000000   \n",
       "\n",
       "      distance1_4   distance2_3  distance2_4  distance3_4  distance_centroid1  \\\n",
       "4602          NaN   1063.893586          NaN          NaN          494.067528   \n",
       "4603          NaN    552.560594          NaN          NaN          550.706261   \n",
       "4604          NaN   4337.165372          NaN          NaN         5930.817269   \n",
       "4605          NaN  36001.700680          NaN          NaN        17388.236060   \n",
       "4606          NaN  85787.088425          NaN          NaN        42190.605451   \n",
       "...           ...           ...          ...          ...                 ...   \n",
       "4774          NaN   -999.000000          NaN          NaN         -999.000000   \n",
       "4775          NaN   -999.000000          NaN          NaN         -999.000000   \n",
       "4776          NaN   -999.000000          NaN          NaN         -999.000000   \n",
       "4777          NaN   -999.000000          NaN          NaN         -999.000000   \n",
       "4778          NaN   -999.000000          NaN          NaN         -999.000000   \n",
       "\n",
       "      distance_centroid2  distance_centroid3  distance_centroid4  cluster_A  \\\n",
       "4602          723.783612          405.121495                 NaN        3.0   \n",
       "4603            0.937820          551.633409                 NaN        3.0   \n",
       "4604         5133.991316          796.825977                 NaN        3.0   \n",
       "4605         9333.418760        26685.649642                 NaN        2.0   \n",
       "4606        21921.440394        63946.745933                 NaN        1.0   \n",
       "...                  ...                 ...                 ...        ...   \n",
       "4774         -999.000000         -999.000000                 NaN        0.0   \n",
       "4775         -999.000000         -999.000000                 NaN        0.0   \n",
       "4776         -999.000000         -999.000000                 NaN        0.0   \n",
       "4777         -999.000000         -999.000000                 NaN        0.0   \n",
       "4778         -999.000000         -999.000000                 NaN        0.0   \n",
       "\n",
       "      cluster_B  cluster_C  cluster_D  \n",
       "4602        0.0        0.0        0.0  \n",
       "4603        0.0        0.0        0.0  \n",
       "4604        0.0        0.0        0.0  \n",
       "4605        1.0        0.0        0.0  \n",
       "4606        1.0        1.0        0.0  \n",
       "...         ...        ...        ...  \n",
       "4774        0.0        0.0        0.0  \n",
       "4775        0.0        0.0        0.0  \n",
       "4776        0.0        0.0        0.0  \n",
       "4777        0.0        0.0        0.0  \n",
       "4778        0.0        0.0        0.0  \n",
       "\n",
       "[177 rows x 18 columns]"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "team_distances[team_distances['teamId'] == 27]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [],
   "source": [
    "# as a final step (that might need to be adjusted) we want to get the categorical distance strategy\n",
    "# we do this by merging the individual distances with a predefined .csv file with the correspondence of strategies\n",
    "def getCollab(df, distance = 10000):\n",
    "    \"\"\"\n",
    "    Get categories of collaborations from distance among players. 'distance' is the threshold distance for collaboration.\n",
    "    Input:\n",
    "        A list with format: [player1, distance1_2, distance1_3, distance1_4, distance2_3, distance2_4, distance3_4]\n",
    "    Outputs:\n",
    "        \"solo-solo\" -> everyone is playing solo\n",
    "        \"solo-duo\" -> two players are together and the other two are solo\n",
    "        \"solo-squad\" -> three players are together and the other one is solo\n",
    "        \"duo-duo\" -> players are in two pairs\n",
    "        \"squad-squad\" -> everyone is together\n",
    "    \"\"\"\n",
    "    def getCatDistance(x, distance = distance):\n",
    "        if x < 0:\n",
    "            return 'd' #dead\n",
    "        if x <= distance:\n",
    "            return 'c' #close\n",
    "        elif x > distance:\n",
    "            return 'f' #far\n",
    "        elif np.isnan(x):\n",
    "            return 'd' #we count extra players in teams of less than 4 people as dead players\n",
    "    \n",
    "    index_player1 = list(df.columns).index('player1')\n",
    "    data = df.iloc[:,index_player1:]\n",
    "    data[['distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4']] = data.iloc[:,1:].applymap(getCatDistance)\n",
    "    categories = pd.read_csv('dist_categories.csv', index_col=0)\n",
    "    transformed_data = pd.merge(data, categories, how='left', \n",
    "                               on = ['player1', 'distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4'])\n",
    "    return transformed_data['categ']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>teamId</th>\n",
       "      <th>player1</th>\n",
       "      <th>distance1_2</th>\n",
       "      <th>distance1_3</th>\n",
       "      <th>distance1_4</th>\n",
       "      <th>distance2_3</th>\n",
       "      <th>distance2_4</th>\n",
       "      <th>distance3_4</th>\n",
       "      <th>categ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>2123.934556</td>\n",
       "      <td>322920.150919</td>\n",
       "      <td>3627.578393</td>\n",
       "      <td>324680.657433</td>\n",
       "      <td>1503.643907</td>\n",
       "      <td>325929.296086</td>\n",
       "      <td>solo-squad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>11.0</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>1810.805346</td>\n",
       "      <td>6604.255870</td>\n",
       "      <td>3147.374302</td>\n",
       "      <td>4793.450740</td>\n",
       "      <td>1336.569115</td>\n",
       "      <td>3456.881629</td>\n",
       "      <td>squad-squad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>21.0</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>1836.793946</td>\n",
       "      <td>3029.650677</td>\n",
       "      <td>3229.368359</td>\n",
       "      <td>4866.444618</td>\n",
       "      <td>1392.574594</td>\n",
       "      <td>6259.018924</td>\n",
       "      <td>squad-squad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>31.0</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>1847.789220</td>\n",
       "      <td>130.431016</td>\n",
       "      <td>3235.395030</td>\n",
       "      <td>1976.990655</td>\n",
       "      <td>1387.605852</td>\n",
       "      <td>3364.558374</td>\n",
       "      <td>squad-squad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>41.0</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>3535.862130</td>\n",
       "      <td>20330.206011</td>\n",
       "      <td>19105.095125</td>\n",
       "      <td>16830.435676</td>\n",
       "      <td>18184.129177</td>\n",
       "      <td>24611.498555</td>\n",
       "      <td>solo-duo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4369</td>\n",
       "      <td>1701.0</td>\n",
       "      <td>27</td>\n",
       "      <td>dead</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>dead-dead</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4370</td>\n",
       "      <td>1711.0</td>\n",
       "      <td>27</td>\n",
       "      <td>dead</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>dead-dead</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4371</td>\n",
       "      <td>1721.0</td>\n",
       "      <td>27</td>\n",
       "      <td>dead</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>dead-dead</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4372</td>\n",
       "      <td>1731.0</td>\n",
       "      <td>27</td>\n",
       "      <td>dead</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>dead-dead</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4373</td>\n",
       "      <td>1741.0</td>\n",
       "      <td>27</td>\n",
       "      <td>dead</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>-999.000000</td>\n",
       "      <td>dead-dead</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4374 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        time  teamId player1  distance1_2    distance1_3   distance1_4  \\\n",
       "0        1.0       1   alive  2123.934556  322920.150919   3627.578393   \n",
       "1       11.0       1   alive  1810.805346    6604.255870   3147.374302   \n",
       "2       21.0       1   alive  1836.793946    3029.650677   3229.368359   \n",
       "3       31.0       1   alive  1847.789220     130.431016   3235.395030   \n",
       "4       41.0       1   alive  3535.862130   20330.206011  19105.095125   \n",
       "...      ...     ...     ...          ...            ...           ...   \n",
       "4369  1701.0      27    dead  -999.000000    -999.000000   -999.000000   \n",
       "4370  1711.0      27    dead  -999.000000    -999.000000   -999.000000   \n",
       "4371  1721.0      27    dead  -999.000000    -999.000000   -999.000000   \n",
       "4372  1731.0      27    dead  -999.000000    -999.000000   -999.000000   \n",
       "4373  1741.0      27    dead  -999.000000    -999.000000   -999.000000   \n",
       "\n",
       "        distance2_3   distance2_4    distance3_4        categ  \n",
       "0     324680.657433   1503.643907  325929.296086   solo-squad  \n",
       "1       4793.450740   1336.569115    3456.881629  squad-squad  \n",
       "2       4866.444618   1392.574594    6259.018924  squad-squad  \n",
       "3       1976.990655   1387.605852    3364.558374  squad-squad  \n",
       "4      16830.435676  18184.129177   24611.498555     solo-duo  \n",
       "...             ...           ...            ...          ...  \n",
       "4369    -999.000000   -999.000000    -999.000000    dead-dead  \n",
       "4370    -999.000000   -999.000000    -999.000000    dead-dead  \n",
       "4371    -999.000000   -999.000000    -999.000000    dead-dead  \n",
       "4372    -999.000000   -999.000000    -999.000000    dead-dead  \n",
       "4373    -999.000000   -999.000000    -999.000000    dead-dead  \n",
       "\n",
       "[4374 rows x 10 columns]"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we apply it and we are done!\n",
    "team_distances['categ'] = getCollab(team_distances)\n",
    "team_distances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get Players Ranking"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This part has been moved to \"ranking_info.ipynb\" file"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## AUTOMATE PROCESS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transformDistances(file):\n",
    "    with open(file, 'r') as f:\n",
    "        json_data = f.read()\n",
    "        f.close()\n",
    "    data = []\n",
    "    for i in json.loads(json_data):\n",
    "        if i[\"_T\"] in (\"LogPlayerPosition\", \"LogParachuteLanding\"):\n",
    "            data.append(i)\n",
    "    # get events data into pandas dataframe\n",
    "    data = pd.read_json(json.dumps(data))\n",
    "    # get player information\n",
    "    from pandas.io.json import json_normalize\n",
    "    data = pd.concat([data, json_normalize(data['character'])], axis=1).drop(['character', 'zone'], axis=1)\n",
    "    data.vehicle = data.vehicle.apply(lambda x: {} if pd.isna(x) else x)\n",
    "    data['vehicle'] = json_normalize(data.vehicle)['vehicleType']\n",
    "    \n",
    "    # harmonize coordinates across time\n",
    "    data_new_coords = data[['elapsedTime','name','location.x','location.y', 'vehicle']].sort_values(['name','elapsedTime']).drop_duplicates()\n",
    "    data_new_coords = data_new_coords.dropna(subset = ['elapsedTime'])\n",
    "    # we round coordinates for ease of calculations, and because a difference in location smaller than a cm is negligible\n",
    "    data_new_coords['location.x'] = data_new_coords['location.x'].apply(round)\n",
    "    data_new_coords['location.y'] = data_new_coords['location.y'].apply(round)\n",
    "    # since different elapsed times don't show the location of all players, we want to interpolate the position of the players at each\n",
    "    # point in time. e.g. if player a is in position (x,y) at second 10.0 and in position (x+10,y+10) at second 20.0\n",
    "    # we can expect that this player will be at position (x+1,y+1) at second 11.0, and so on\n",
    "    def getNewCoord(t, n):\n",
    "        aa = data_new_coords[(data_new_coords['elapsedTime'] < t) & (data_new_coords['name'] == n)] # get previous observations\n",
    "        bb = data_new_coords[(data_new_coords['elapsedTime'] > t) & (data_new_coords['name'] == n)] # get following observations\n",
    "        if (bool(aa['location.x'].any())) & (bool(bb['location.x'].any())): # if we have both previous and following observation\n",
    "            a = aa.iloc[-1] # get last previous observation\n",
    "            b = bb.iloc[0] # get first following observation\n",
    "            time = b[0] - a[0] # calculate the time difference\n",
    "            # get x and y coordinate by calculating movement speed in cm/s and get full coordinate\n",
    "            new_coord_x = a[2] + ((b[2] - a[2]) / time)*(t-a[0])  \n",
    "            new_coord_y = a[3] + ((b[3] - a[3]) / time)*(t-a[0])\n",
    "            in_aircraft = b['vehicle']\n",
    "            return new_coord_x, new_coord_y, in_aircraft\n",
    "        elif (not bool(aa['location.x'].any())) & (bool(bb['location.x'].any())):\n",
    "            # if there is no previous observation, set first following location to be current location\n",
    "            in_aircraft = bb['vehicle']\n",
    "            return bb.iat[0,2], bb.iat[0,3], in_aircraft\n",
    "\n",
    "    # MUST FIX THIS PART, LOCATION EVENT KEEPS APPEARING FOR SOME TIME AFTER THEY DIE\n",
    "    # LOOK AT HEALTH DATA\n",
    "        elif (bool(aa['location.x'].any())) & (not bool(bb['location.x'].any())):\n",
    "            # if there is no following observation, we assume the player has died (we hope that only in the match)\n",
    "            return 'dead', 'dead', 'dead'\n",
    "\n",
    "    names=[]\n",
    "    times=[]\n",
    "    x_coords=[]\n",
    "    y_coords=[]\n",
    "    aircraft=[]\n",
    "    # we will get the location every 10 seconds from second 1 (e.g. 1.0, 11.0, 21.0), and at the last second\n",
    "    times_list = data_new_coords.elapsedTime.unique().astype(int)\n",
    "    unique_times_10s = np.arange(1, max(times_list), 10).tolist()\n",
    "    unique_times_10s.append(max(times_list))\n",
    "    for t in unique_times_10s:\n",
    "        for n in data_new_coords.name.unique():\n",
    "            a = data_new_coords[(data_new_coords['elapsedTime'] == t) & (data_new_coords['name'] == n)]\n",
    "            if a['location.x'].any():\n",
    "                names.append(n)\n",
    "                times.append(t)\n",
    "                x_coords.append(a.iat[0,2])\n",
    "                y_coords.append(a.iat[0,3])\n",
    "                aircraft.append(a['vehicle'])\n",
    "            else:\n",
    "                x_coord, y_coord, in_aircraft = getNewCoord(t,n)\n",
    "                names.append(n)\n",
    "                times.append(t)\n",
    "                x_coords.append(x_coord)\n",
    "                y_coords.append(y_coord)\n",
    "                aircraft.append(in_aircraft)\n",
    "    players_location = pd.DataFrame(list(zip(times,names,x_coords, y_coords, aircraft)),\n",
    "                 columns=['time','name','location.x','location.y', 'inAircraft'])\n",
    "    players_location['inAircraft'] = players_location['inAircraft'].apply(lambda x: 1 if str(x) == 'TransportAircraft' else 0)\n",
    "    \n",
    "    # encode player names\n",
    "    players = data[['name','accountId', 'teamId']].drop_duplicates()\n",
    "    players = players.sort_values('teamId').reset_index().drop(['index'], axis=1)\n",
    "    players_location = pd.merge(players_location,\n",
    "                            players,\n",
    "                            left_on = 'name',\n",
    "                            right_on = 'name',\n",
    "                            how = 'left')\n",
    "    players_location = players_location[players_location['time'] >= 0].sort_values(['teamId','time', 'name']).reset_index().drop(['index'], axis=1)\n",
    "    \n",
    "    # get centroids\n",
    "    def getCentroid(*args):\n",
    "        \"Returns a single coordinate centroid\"\n",
    "        locs = [i for i in [*args[0]] if type(i) != str]\n",
    "        s = sum(locs)\n",
    "        n = len(locs)\n",
    "        if n == 0:\n",
    "            return 0\n",
    "        else:\n",
    "            return s/n\n",
    "    teams = []\n",
    "    times = []\n",
    "    centroidsx = []\n",
    "    centroidsy = []\n",
    "    for teamId in players_location.teamId.unique():\n",
    "        for time in players_location.time.unique():\n",
    "            df = players_location[(players_location.teamId == teamId) & (players_location.time == time)]\n",
    "            centroidx = getCentroid(df['location.x'])\n",
    "            centroidy = getCentroid(df['location.y'])\n",
    "            teams.append(teamId)\n",
    "            times.append(time)\n",
    "            centroidsx.append(centroidx)\n",
    "            centroidsy.append(centroidy)\n",
    "    dict = {'teamId':teams, 'time':times, 'centroid.x':centroidsx, 'centroid.y':centroidsy}\n",
    "    players_location = pd.merge(players_location, pd.DataFrame(dict), how = 'left', on = ['teamId','time'])\n",
    "        \n",
    "    # get distance to centroid\n",
    "    def getDistance(locx_1, locy_1, locx_2, locy_2):\n",
    "        \"\"\"\n",
    "        Returns the euclidean distance between two points in a plane given their two coordinates, x and y\n",
    "        \"\"\"\n",
    "        try:\n",
    "            return ((locx_2 - locx_1)**2 + (locy_2 - locy_1)**2)**(1/2)\n",
    "        # if player is dead ('dead' in the location), we return a negative distance\n",
    "        except:\n",
    "            return -999\n",
    "    players_location['distance_centroid'] = np.vectorize(getDistance)(players_location['location.x'], players_location['location.y'], players_location['centroid.x'], players_location['centroid.y'])\n",
    "    \n",
    "    # get distance between pairs of players\n",
    "    distances = np.empty((0,4))\n",
    "    for team in players_location.teamId.unique():\n",
    "        df1 = players_location[players_location.teamId == team]\n",
    "        players = {}\n",
    "        # I have to check why I did this in the first place because I think it might not be useful anymore\n",
    "        for i, player in zip(range(len(data.name.unique())), data.name.unique()):\n",
    "            players[f'player{i+1}'] = player\n",
    "        for time in df1.time.unique():\n",
    "            df2 = df1[df1['time'] == time]\n",
    "            if (len(df2) > 4) and (time==1):\n",
    "                name = df2[df2.applymap(type)['location.x'] == str]['name'].iloc[0]\n",
    "            if len(df2) > 4:\n",
    "                df2 = df2[df2['name'] != name]\n",
    "            # get a diagonal matrix with the distances between every pair of players from the same team in a time\n",
    "            # including themselves\n",
    "            d = np.array([[getDistance(i,j,k,l) for i,j in zip(df2['location.x'], df2['location.y'])] \n",
    "                          for k,l in zip(df2['location.x'], df2['location.y'])])\n",
    "            if d.shape[1] == 4:\n",
    "                distances = np.concatenate((distances, d))\n",
    "            # if the team has less than 4 players, we add a Null value for the column of the non-existent player\n",
    "            # this is just for the purpose of maintaining the shape of the matrix\n",
    "            else:\n",
    "                h = np.full((d.shape[1], 4 - d.shape[1]), np.nan)\n",
    "                d = np.column_stack((d, h))\n",
    "                distances = np.concatenate((distances, d))\n",
    "    players_location = pd.concat((players_location, pd.DataFrame(distances, columns = ['player1','player2','player3','player4'])),\n",
    "                            axis = 1)\n",
    "\n",
    "    # getting clusters of players\n",
    "    times = []\n",
    "    names = []\n",
    "    clusters = []\n",
    "    for teamId in players_location.teamId.unique():\n",
    "        for time in players_location.time.unique():\n",
    "            df = players_location[(players_location.teamId == teamId) & (players_location.time == time)]\n",
    "            X = df[['location.x','location.y']]\n",
    "            X = X.applymap(lambda x: -999 if x == 'dead' else int(x))\n",
    "            clustering = DBSCAN(eps=10000, min_samples=1).fit(X)\n",
    "            times = np.concatenate((times, df['time'].tolist())).tolist()\n",
    "            names = np.concatenate((names, df['name'].tolist())).tolist()\n",
    "            clusters = np.concatenate((clusters, clustering.labels_.tolist())).tolist()\n",
    "    players_location = pd.merge(players_location, pd.DataFrame(list(zip(times,names,clusters)), columns = ['time','name','cluster']),\n",
    "                            how = 'left', on = ['time','name'])\n",
    "    players_location.loc[players_location['location.x'] == 'dead', 'cluster'] = -1\n",
    "    l_clusters = {0.0:'A', 1.0:'B', 2.0:'C', 3.0:'D', -1.0:'Z'}\n",
    "    players_location.cluster = players_location.cluster.map(l_clusters)\n",
    "    cluster_count = players_location.pivot_table(values = ['name'], index = ['time', 'teamId'], columns=['cluster'], aggfunc = 'count')\n",
    "    cluster_count.columns = cluster_count.columns.droplevel()\n",
    "    cluster_count = cluster_count.reset_index()\n",
    "    cluster_count = cluster_count.fillna(0)\n",
    "    cluster_count\n",
    "\n",
    "    # get the distances in flat format for each team\n",
    "    rows = []\n",
    "    matchid = json.loads(json_data)[0][\"MatchId\"]\n",
    "    matchid = matchid.split('.')[-1]\n",
    "    for teamId in players_location.teamId.unique():\n",
    "        n_players = len(players_location[players_location['teamId'] == teamId].name.unique())\n",
    "        for time in players_location.time.unique():\n",
    "            df = players_location[(players_location['teamId'] == teamId) & (players_location['time'] == time)]\n",
    "            alive_players = len([x for x in df['location.x'] if type(x) != str])\n",
    "            in_aircraft = sum(df['inAircraft'])\n",
    "            if df.shape[0] == 4:\n",
    "                row = (matchid, time, teamId, n_players, alive_players, in_aircraft, \n",
    "                       df.iat[0,3], df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], df.iat[2,13],\n",
    "                       df.iat[0,9], df.iat[1,9], df.iat[2,9], df.iat[3,9])\n",
    "            # teams with less than 3 players don't have a third row in their inner matrixes, so we manually input a null\n",
    "            elif df.shape[0] == 3:\n",
    "                row = (matchid, time, teamId, n_players, alive_players, in_aircraft, \n",
    "                       df.iat[0,3], df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], df.iat[2,13],\n",
    "                       df.iat[0,9], df.iat[1,9], df.iat[2,9], np.nan)            \n",
    "            elif df.shape[0] == 2:\n",
    "                row = (matchid, time, teamId, n_players, alive_players, in_aircraft, \n",
    "                       df.iat[0,3], df.iat[0,11], df.iat[0,12], df.iat[0,13], df.iat[1,12], df.iat[1,13], np.nan,\n",
    "                       df.iat[0,9], df.iat[1,9], np.nan, np.nan)\n",
    "            else:\n",
    "                row = (matchid, time, teamId, n_players, alive_players, in_aircraft, \n",
    "                       df.iat[0,3], df.iat[0,11], df.iat[0,12], df.iat[0,13], np.nan, np.nan, np.nan,\n",
    "                       df.iat[0,9], np.nan, np.nan, np.nan)\n",
    "            rows.append(row)\n",
    "    columns = ['matchId', 'time', 'teamId', 'n_players', 'n_alive', 'in_aircraft', 'player1',\n",
    "               'distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4',\n",
    "               'distance_centroid1', 'distance_centroid2', 'distance_centroid3', 'distance_centroid4']\n",
    "    team_distances = pd.DataFrame(rows, columns=columns)\n",
    "    team_distances['player1'] = ['alive' if type(x) != str else 'dead' for x in team_distances['player1']]\n",
    "    team_distances = pd.merge(team_distances, cluster_count, how='left', on=['time','teamId']).drop('Z', axis=1)\n",
    "    team_distances = team_distances.rename(columns = {'A': 'cluster_A', 'B': 'cluster_B', 'C': 'cluster_C', 'D': 'cluster_D'})\n",
    "    \n",
    "    # get categorical collaboration\n",
    "    def getCollab(df, distance = 10000):\n",
    "        \"\"\"\n",
    "        Get categories of collaborations from distance among players. 'distance' is the threshold distance for collaboration.\n",
    "        Input:\n",
    "            A list with format: [player1, distance1_2, distance1_3, distance1_4, distance2_3, distance2_4, distance3_4]\n",
    "        Outputs:\n",
    "            \"solo-solo\" -> everyone is playing solo\n",
    "            \"solo-duo\" -> two players are together and the other two are solo\n",
    "            \"solo-squad\" -> three players are together and the other one is solo\n",
    "            \"duo-duo\" -> players are in two pairs\n",
    "            \"squad-squad\" -> everyone is together\n",
    "        \"\"\"\n",
    "        def getCatDistance(x, distance = distance):\n",
    "            if x < 0:\n",
    "                return 'd' #dead\n",
    "            if x <= distance:\n",
    "                return 'c' #close\n",
    "            elif x > distance:\n",
    "                return 'f' #far\n",
    "            elif np.isnan(x):\n",
    "                return 'd' #we count extra players in teams of less than 4 people as dead players\n",
    "\n",
    "        index_player1 = list(df.columns).index('player1')\n",
    "        index_centroid1 = list(df.columns).index('distance_centroid1')\n",
    "        data = df.iloc[:,index_player1:index_centroid1]\n",
    "        data[['distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4']] = data.iloc[:,1:].applymap(getCatDistance)\n",
    "        categories = pd.read_csv('dist_categories.csv', index_col=0)\n",
    "        transformed_data = pd.merge(data, categories, how='left', \n",
    "                                   on = ['player1', 'distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4'])\n",
    "        transformed_data['strategy'] = data.apply(lambda x: f'{x[1]}{x[2]}{x[3]}{x[4]}{x[5]}{x[6]}', axis=1)\n",
    "        return transformed_data[['categ', 'strategy']]\n",
    "    \n",
    "    team_distances[['categ', 'strategy']] = getCollab(team_distances)\n",
    "    \n",
    "    return team_distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'telemetry_data/2020-04-27/console/01ece79a-8fe8-4fc7-824e-da292ab9e51d.json'"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ranking = pd.read_csv('ranking_data/pc/2020-04-22.csv', index_col = 0)\n",
    "failed = []\n",
    "for filename in os.listdir('telemetry_data/2020-04-22/console'):\n",
    "    file = f'telemetry_data/2020-04-22/console/{filename}'\n",
    "    try:\n",
    "        df = transformDistances(file)\n",
    "        dfn = pd.merge(df, ranking, how='left', on=['matchId', 'teamId'])\n",
    "        dfn = dfn.drop('player1', axis=1)\n",
    "        newname = filename.split('.')[0]\n",
    "        if 'cluster_D' not in dfn.columns:\n",
    "            dfn['cluster_D'] = 0.0\n",
    "        if 'cluster_C' not in dfn.columns:\n",
    "            dfn['cluster_C'] = 0.0\n",
    "        if 'cluster_B' not in dfn.columns:\n",
    "            dfn['cluster_B'] = 0.0\n",
    "        columns = ['matchId', 'time', 'teamId', 'n_players', 'n_alive', 'in_aircraft',\n",
    "                    'distance1_2', 'distance1_3', 'distance1_4', 'distance2_3', 'distance2_4', 'distance3_4',\n",
    "                    'distance_centroid1', 'distance_centroid2', 'distance_centroid3', 'distance_centroid4',\n",
    "                    'cluster_A', 'cluster_B', 'cluster_C', 'cluster_D', 'categ', 'strategy', 'ranking']\n",
    "        dfn = dfn[columns]\n",
    "        dfn.to_csv(path_or_buf=f'telemetry_data/clean/{newname}.csv')\n",
    "        print(f'{filename} done')\n",
    "    except (KeyboardInterrupt, SystemExit):\n",
    "        raise\n",
    "    except Exception as e:\n",
    "        newname = filename.split('.')[0]\n",
    "        failed.append(newname)\n",
    "        print(f'!!! {filename} FAILED, {type(e)}:{e}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "failed = list(set(failed))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "606"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len([i for i,x in enumerate(os.listdir('telemetry_data/2020-04-27/console'))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>matchId</th>\n",
       "      <th>time</th>\n",
       "      <th>teamId</th>\n",
       "      <th>n_players</th>\n",
       "      <th>player1</th>\n",
       "      <th>distance1_2</th>\n",
       "      <th>distance1_3</th>\n",
       "      <th>distance1_4</th>\n",
       "      <th>distance2_3</th>\n",
       "      <th>distance2_4</th>\n",
       "      <th>distance3_4</th>\n",
       "      <th>categ</th>\n",
       "      <th>ranking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0ae75e5f-1897-455c-9ab3-0dc83c226c3a</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>dead-solo</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0ae75e5f-1897-455c-9ab3-0dc83c226c3a</td>\n",
       "      <td>11.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>dead-solo</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0ae75e5f-1897-455c-9ab3-0dc83c226c3a</td>\n",
       "      <td>21.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>dead-solo</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0ae75e5f-1897-455c-9ab3-0dc83c226c3a</td>\n",
       "      <td>31.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>dead-solo</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0ae75e5f-1897-455c-9ab3-0dc83c226c3a</td>\n",
       "      <td>41.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>alive</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>dead-solo</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                matchId  time  teamId  n_players player1  \\\n",
       "0  0ae75e5f-1897-455c-9ab3-0dc83c226c3a   1.0       1          1   alive   \n",
       "1  0ae75e5f-1897-455c-9ab3-0dc83c226c3a  11.0       1          1   alive   \n",
       "2  0ae75e5f-1897-455c-9ab3-0dc83c226c3a  21.0       1          1   alive   \n",
       "3  0ae75e5f-1897-455c-9ab3-0dc83c226c3a  31.0       1          1   alive   \n",
       "4  0ae75e5f-1897-455c-9ab3-0dc83c226c3a  41.0       1          1   alive   \n",
       "\n",
       "   distance1_2  distance1_3  distance1_4  distance2_3  distance2_4  \\\n",
       "0          NaN          NaN          NaN          NaN          NaN   \n",
       "1          NaN          NaN          NaN          NaN          NaN   \n",
       "2          NaN          NaN          NaN          NaN          NaN   \n",
       "3          NaN          NaN          NaN          NaN          NaN   \n",
       "4          NaN          NaN          NaN          NaN          NaN   \n",
       "\n",
       "   distance3_4      categ  ranking  \n",
       "0          NaN  dead-solo       27  \n",
       "1          NaN  dead-solo       27  \n",
       "2          NaN  dead-solo       27  \n",
       "3          NaN  dead-solo       27  \n",
       "4          NaN  dead-solo       27  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ranking = pd.read_csv('ranking_data/console/2020-04-27.csv', index_col = 0)\n",
    "dfm = pd.merge(df, ranking, how='left', on=['matchId', 'teamId'])\n",
    "dfm.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Graphing collaboration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = dfm.copy()\n",
    "a = a.sort_values(by=['time','categ'], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = a['time']\n",
    "y = a['categ']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (14,8))\n",
    "#plt.scatter(a.time, a.categ, c=a.ranking, cmap='viridis')\n",
    "for rank in a.ranking.unique():\n",
    "    if rank < 6:\n",
    "        plt.plot(a.time[a.ranking == rank], a.categ[a.ranking == rank], c = 'green')\n",
    "    else:\n",
    "        plt.plot(a.time[a.ranking == rank], a.categ[a.ranking == rank], c = 'blue', alpha=0.10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Export to .csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "# export match telemetry data into .csv format\n",
    "data.to_csv(path_or_buf=f'{matchid}.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### EXTRA: Check distances distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.00         0.000000\n",
       "0.05       417.202108\n",
       "0.10       803.991275\n",
       "0.15      1275.678684\n",
       "0.20      1903.554244\n",
       "0.25      2652.179041\n",
       "0.30      3514.965001\n",
       "0.35      4620.385915\n",
       "0.40      5917.666320\n",
       "0.45      7617.327949\n",
       "0.50      9520.768264\n",
       "0.55     11565.626537\n",
       "0.60     14577.902657\n",
       "0.65     18439.278686\n",
       "0.70     23214.105653\n",
       "0.75     29108.496210\n",
       "0.80     36950.220271\n",
       "0.85     46505.286428\n",
       "0.90     63428.487948\n",
       "0.95    139534.596795\n",
       "dtype: float64"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances = pd.concat([team_distances['distance1_2'],team_distances['distance1_3'],team_distances['distance1_4'],\n",
    "                     team_distances['distance2_3'],team_distances['distance2_4'],team_distances['distance3_4']])\n",
    "distances = distances[distances >=0]\n",
    "hist = distances.hist(bins=50, figsize=(14,8))\n",
    "distances.quantile(np.arange(0.0,1.0,0.05))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### EXTRA: Check how fast players move"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get coordinates as points in a plane (only x and y coordinates)\n",
    "data_sorted = data[['elapsedTime','name','location.x','location.y']].sort_values(['name','elapsedTime'])\n",
    "data_sorted = data_sorted.drop_duplicates().reset_index().drop('index', axis=1)\n",
    "data_sorted['coord_t'] = data_sorted.apply(lambda row: [row['location.x'], row['location.y']], axis=1)\n",
    "data_sorted['elapsedTime_1'] = data_sorted['elapsedTime'].shift(1)\n",
    "data_sorted['coord_1'] = data_sorted['coord_t'].shift(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>elapsedTime</th>\n",
       "      <th>name</th>\n",
       "      <th>location.x</th>\n",
       "      <th>location.y</th>\n",
       "      <th>coord_t</th>\n",
       "      <th>elapsedTime_1</th>\n",
       "      <th>coord_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10Baerini</td>\n",
       "      <td>719535.500000</td>\n",
       "      <td>546555.8750</td>\n",
       "      <td>[719535.5, 546555.875]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10Baerini</td>\n",
       "      <td>254098.703125</td>\n",
       "      <td>954479.7500</td>\n",
       "      <td>[254098.703125, 954479.75]</td>\n",
       "      <td>0.0</td>\n",
       "      <td>[719535.5, 546555.875]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10Baerini</td>\n",
       "      <td>258307.750000</td>\n",
       "      <td>811171.7500</td>\n",
       "      <td>[258307.75, 811171.75]</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[254098.703125, 954479.75]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>21.0</td>\n",
       "      <td>10Baerini</td>\n",
       "      <td>262448.937500</td>\n",
       "      <td>670174.2500</td>\n",
       "      <td>[262448.9375, 670174.25]</td>\n",
       "      <td>11.0</td>\n",
       "      <td>[258307.75, 811171.75]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>30.0</td>\n",
       "      <td>10Baerini</td>\n",
       "      <td>266590.593750</td>\n",
       "      <td>529160.6875</td>\n",
       "      <td>[266590.59375, 529160.6875]</td>\n",
       "      <td>21.0</td>\n",
       "      <td>[262448.9375, 670174.25]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   elapsedTime       name     location.x   location.y  \\\n",
       "0          0.0  10Baerini  719535.500000  546555.8750   \n",
       "1          1.0  10Baerini  254098.703125  954479.7500   \n",
       "2         11.0  10Baerini  258307.750000  811171.7500   \n",
       "3         21.0  10Baerini  262448.937500  670174.2500   \n",
       "4         30.0  10Baerini  266590.593750  529160.6875   \n",
       "\n",
       "                       coord_t  elapsedTime_1                     coord_1  \n",
       "0       [719535.5, 546555.875]            NaN                         NaN  \n",
       "1   [254098.703125, 954479.75]            0.0      [719535.5, 546555.875]  \n",
       "2       [258307.75, 811171.75]            1.0  [254098.703125, 954479.75]  \n",
       "3     [262448.9375, 670174.25]           11.0      [258307.75, 811171.75]  \n",
       "4  [266590.59375, 529160.6875]           21.0    [262448.9375, 670174.25]  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# eliminate the first observation in coord_t-1 for each player\n",
    "coord_t1 = []\n",
    "elapsedTime = []\n",
    "for i in data_sorted.name.unique():\n",
    "    a = data_sorted[data_sorted['name'] == i]['coord_1'].tolist()\n",
    "    b = data_sorted[data_sorted['name'] == i]['elapsedTime_1'].tolist()\n",
    "    a[0] = np.nan\n",
    "    b[0] = np.nan\n",
    "    coord_t1 = coord_t1 + a\n",
    "    elapsedTime = elapsedTime + b\n",
    "\n",
    "data_sorted['coord_1'] = coord_t1\n",
    "data_sorted['elapsedTime_1'] = elapsedTime\n",
    "data_sorted.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get the movement speed\n",
    "speed = []\n",
    "for i,j,k,l in zip(data_sorted.coord_t, data_sorted.coord_1, data_sorted.elapsedTime, data_sorted.elapsedTime_1):\n",
    "    if type(j) == float:\n",
    "        speed.append(np.nan)\n",
    "    elif (k-l) == 0:\n",
    "        speed.append(999999.9)\n",
    "    else:\n",
    "        speed.append((((i[0] - j[0])**2 + (i[1] - j[1])**2)**(1/2)) / (k - l))\n",
    "data_sorted['speed'] = speed\n",
    "data_sorted_speed = data_sorted.dropna(subset=['speed'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x213a17e6888>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_sorted_speed[data_sorted_speed['speed']<999999]['speed'].hist(bins=150, figsize=[18,8])"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}