{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import math\n",
    "import time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Table of contents\n",
    "1. [Loading data from the website](#websiteData)\n",
    "2. [Loading data locally](#localData)\n",
    "3. [Dataset creation initial functions](#datasetCreate)  \n",
    "    a. [Victory Point (target) dataset](#VPData)  \n",
    "    b. [Features dataset](#featuresData)\n",
    "4. [Speeding check & speed up dataset creation & sense-checking on small batch](#speedUp)\n",
    "5. [Cleaning & filtering the data](#cleanUp)\n",
    "6. [Creating the full dataset](#fullDataset)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading data from the website (not using this right now - for now we'll just load locally) <a name=\"websiteData\"></a>\n",
    "But we'll keep this section here for now, commented out. The reason we're not using this is because the jsons are nested, which would require a bit of time to unpack. I'd like to focus on the cleaning / making some form of useful data and modelling, for now. Later we can go back and get more data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# bigoldlist = ['2013-02.json',\n",
    "# '2013-03.json',\n",
    "# '2013-04.json',\n",
    "# '2013-05.json',\n",
    "# '2013-06.json',\n",
    "# '2013-07.json',\n",
    "# '2013-08.json',\n",
    "# '2013-09.json',\n",
    "# '2013-10.json',\n",
    "# '2013-11.json',\n",
    "# '2013-12.json',\n",
    "# '2014-01.json',\n",
    "# '2014-02.json',\n",
    "# '2014-03.json',\n",
    "# '2014-04.json',\n",
    "# '2014-05.json',\n",
    "# '2014-06.json',\n",
    "# '2014-07.json',\n",
    "# '2014-08.json',\n",
    "# '2014-09.json',\n",
    "# '2014-10.json',\n",
    "# '2014-11.json',\n",
    "# '2014-12.json',\n",
    "# '2015-01.json',\n",
    "# '2015-02.json',\n",
    "# '2015-03.json',\n",
    "# '2015-04.json',\n",
    "# '2015-05.json',\n",
    "# '2015-06.json',\n",
    "# '2015-07.json',\n",
    "# '2015-08.json',\n",
    "# '2015-09.json',\n",
    "# '2015-10.json',\n",
    "# '2015-11.json',\n",
    "# '2015-12.json',\n",
    "# '2016-01.json',\n",
    "# '2016-02.json',\n",
    "# '2016-03.json',\n",
    "# '2016-04.json',\n",
    "# '2016-05.json',\n",
    "# '2016-06.json',\n",
    "# '2016-07.json',\n",
    "# '2016-08.json',\n",
    "# '2016-09.json',\n",
    "# '2016-10.json',\n",
    "# '2016-11.json',\n",
    "# '2016-12.json',\n",
    "# '2017-01.json',\n",
    "# '2017-02.json',\n",
    "# '2017-03.json',\n",
    "# '2017-04.json',\n",
    "# '2017-05.json',\n",
    "# '2017-06.json',\n",
    "# '2017-07.json',\n",
    "# '2017-08.json',\n",
    "# '2017-09.json',\n",
    "# '2017-10.json',\n",
    "# '2017-11.json',\n",
    "# '2017-12.json',\n",
    "# '2018-01.json',\n",
    "# '2018-02.json',\n",
    "# '2018-03.json',\n",
    "# '2018-04.json',\n",
    "# '2018-05.json',\n",
    "# '2018-06.json',\n",
    "# '2018-07.json',\n",
    "# '2018-08.json',\n",
    "# '2018-09.json',\n",
    "# '2018-10.json',\n",
    "# '2018-11.json',\n",
    "# '2018-12.json',\n",
    "# '2019-01.json',\n",
    "# '2019-02.json',\n",
    "# '2019-03.json',\n",
    "# '2019-04.json',\n",
    "# '2019-05.json',\n",
    "# '2019-06.json',\n",
    "# '2019-07.json',\n",
    "# '2019-08.json',\n",
    "# '2019-09.json',\n",
    "# '2019-10.json',\n",
    "# '2019-11.json',\n",
    "# '2019-12.json',\n",
    "# '2020-01.json',\n",
    "# '2020-02.json',\n",
    "# '2020-03.json',\n",
    "# '2020-04.json',\n",
    "# '2020-05.json',\n",
    "# '2020-06.json',\n",
    "# '2020-07.json',\n",
    "# '2020-08.json',\n",
    "# '2020-09.json',\n",
    "# '2020-10.json',\n",
    "# '2020-11.json',\n",
    "# '2020-12.json',\n",
    "# '2021-01.json',\n",
    "# '2021-02.json',\n",
    "# '2021-03.json',\n",
    "# '2021-04.json']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Test start of web pipeline (not used)\n",
    "# url = 'https://terra.snellman.net/data/events/' + bigoldlist[0]\n",
    "# eventsdata1 = pd.read_json(url)\n",
    "# print(len(eventsdata1))\n",
    "# eventsdata1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Load in the data remotely\n",
    "# for gameset in bigoldlist:\n",
    "#     url = 'https://terra.snellman.net/data/events/' + gameset\n",
    "#     eventsdata = pd.read_json(url)\n",
    "    \n",
    "#     for eachgame in range(len(gameset)):\n",
    "#         singlerow = eventsdata.iloc[[eachgame]]\n",
    "#         newcol = get_vp_from_game(gameEventRow)\n",
    "#         vpdf = vpdf.append(newdf, ignore_index=True)      "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading in the data Locally <a name=\"localData\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "folderlocation = \"D:/PycharmProjects/TerraBot/terra-mystica\"\n",
    "gameevents = pd.read_csv(f'{folderlocation}/game_events.csv')\n",
    "games = pd.read_csv(f'{folderlocation}/games.csv')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of game events: 73419074\n",
      "Number of games in gameevents: 115612\n",
      "len of games: 115612\n"
     ]
    }
   ],
   "source": [
    "# get some more stats\n",
    "gameslist = list(pd.unique(gameevents['game']))\n",
    "print(f'Number of game events: {len(gameevents)}')\n",
    "print(f\"Number of games in gameevents: {len(gameslist)}\")\n",
    "print(f'len of games: {len(games)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>base_map</th>\n",
       "      <th>game</th>\n",
       "      <th>last_update</th>\n",
       "      <th>player_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>2015-07-22 05:15:51</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>050615</td>\n",
       "      <td>2015-07-13 09:45:00</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>0512</td>\n",
       "      <td>2015-07-03 04:17:31</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>051501</td>\n",
       "      <td>2015-07-16 21:00:54</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>95a66999127893f5925a5f591d54f8bcb9a670e6</td>\n",
       "      <td>060303</td>\n",
       "      <td>2015-07-02 05:20:19</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115607</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>Youreacrookcaptianhook</td>\n",
       "      <td>2014-09-04 05:04:45</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115608</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>YourMomIsAChaosMagician</td>\n",
       "      <td>2014-09-02 00:23:43</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115609</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>ytuwertqwtr</td>\n",
       "      <td>2014-09-28 18:12:35</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115610</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>yuertyqert</td>\n",
       "      <td>2014-09-27 19:01:17</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115611</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>2014-09-10 20:35:32</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>115612 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        base_map                     game  \\\n",
       "0       126fe960806d587c78546b30f1a90853b1ada468               0000000001   \n",
       "1       be8f6ebf549404d015547152d5f2a1906ae8dd90                   050615   \n",
       "2       126fe960806d587c78546b30f1a90853b1ada468                     0512   \n",
       "3       126fe960806d587c78546b30f1a90853b1ada468                   051501   \n",
       "4       95a66999127893f5925a5f591d54f8bcb9a670e6                   060303   \n",
       "...                                          ...                      ...   \n",
       "115607  126fe960806d587c78546b30f1a90853b1ada468   Youreacrookcaptianhook   \n",
       "115608  126fe960806d587c78546b30f1a90853b1ada468  YourMomIsAChaosMagician   \n",
       "115609  126fe960806d587c78546b30f1a90853b1ada468              ytuwertqwtr   \n",
       "115610  126fe960806d587c78546b30f1a90853b1ada468               yuertyqert   \n",
       "115611  126fe960806d587c78546b30f1a90853b1ada468        ZeicheMasZeuchnis   \n",
       "\n",
       "                last_update  player_count  \n",
       "0       2015-07-22 05:15:51             2  \n",
       "1       2015-07-13 09:45:00             4  \n",
       "2       2015-07-03 04:17:31             4  \n",
       "3       2015-07-16 21:00:54             5  \n",
       "4       2015-07-02 05:20:19             4  \n",
       "...                     ...           ...  \n",
       "115607  2014-09-04 05:04:45             3  \n",
       "115608  2014-09-02 00:23:43             4  \n",
       "115609  2014-09-28 18:12:35             2  \n",
       "115610  2014-09-27 19:01:17             2  \n",
       "115611  2014-09-10 20:35:32             3  \n",
       "\n",
       "[115612 rows x 4 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "games"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>event</th>\n",
       "      <th>faction</th>\n",
       "      <th>game</th>\n",
       "      <th>num</th>\n",
       "      <th>round</th>\n",
       "      <th>turn</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>upgrade:SH</td>\n",
       "      <td>halflings</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>order:2</td>\n",
       "      <td>halflings</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>order:2</td>\n",
       "      <td>halflings</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>order:1</td>\n",
       "      <td>halflings</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>order:1</td>\n",
       "      <td>halflings</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73419069</th>\n",
       "      <td>burn</td>\n",
       "      <td>darklings</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73419070</th>\n",
       "      <td>burn</td>\n",
       "      <td>darklings</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73419071</th>\n",
       "      <td>favor:FAV10</td>\n",
       "      <td>darklings</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73419072</th>\n",
       "      <td>town:TW3</td>\n",
       "      <td>darklings</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73419073</th>\n",
       "      <td>action:ACT3</td>\n",
       "      <td>darklings</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>73419074 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                event    faction               game  num  round  turn\n",
       "0          upgrade:SH  halflings         0000000001    1      2     1\n",
       "1             order:2  halflings         0000000001    1      6     1\n",
       "2             order:2  halflings         0000000001    1      2     1\n",
       "3             order:1  halflings         0000000001    1      1     1\n",
       "4             order:1  halflings         0000000001    1      5     1\n",
       "...               ...        ...                ...  ...    ...   ...\n",
       "73419069         burn  darklings  ZeicheMasZeuchnis    4      1     1\n",
       "73419070         burn  darklings  ZeicheMasZeuchnis    3      5     2\n",
       "73419071  favor:FAV10  darklings  ZeicheMasZeuchnis    1      3     3\n",
       "73419072     town:TW3  darklings  ZeicheMasZeuchnis    1      6     2\n",
       "73419073  action:ACT3  darklings  ZeicheMasZeuchnis    1      1     1\n",
       "\n",
       "[73419074 rows x 6 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gameevents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['halflings', 'all', 'mermaids', 'dragonlords', 'riverwalkers',\n",
       "       'yetis', 'darklings', 'chaosmagicians', 'engineers', 'icemaidens',\n",
       "       'nomads', 'auren', 'fakirs', 'witches', 'cultists', 'alchemists',\n",
       "       'swarmlings', 'dwarves', 'shapeshifters', 'nofaction1',\n",
       "       'nofaction4', 'acolytes', 'giants', 'nofaction3', 'nofaction2',\n",
       "       'nofaction5', 'nofaction7'], dtype=object)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "allfactions = pd.unique(gameevents['faction'])\n",
    "allfactions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "gamescoringtiles = pd.read_csv(f'{folderlocation}/game_scoring_tiles.csv')\n",
    "gameoptions = pd.read_csv(f'{folderlocation}/game_options.csv')\n",
    "stats = pd.read_csv(f'{folderlocation}/stats.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>round</th>\n",
       "      <th>tile</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>3</td>\n",
       "      <td>SCORE1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>6</td>\n",
       "      <td>SCORE5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>5</td>\n",
       "      <td>SCORE4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>1</td>\n",
       "      <td>SCORE8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>2</td>\n",
       "      <td>SCORE3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693655</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>4</td>\n",
       "      <td>SCORE1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693656</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>6</td>\n",
       "      <td>SCORE5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693657</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>5</td>\n",
       "      <td>SCORE7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693658</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>1</td>\n",
       "      <td>SCORE3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693659</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>2</td>\n",
       "      <td>SCORE8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>693660 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     game  round    tile\n",
       "0              0000000001      3  SCORE1\n",
       "1              0000000001      6  SCORE5\n",
       "2              0000000001      5  SCORE4\n",
       "3              0000000001      1  SCORE8\n",
       "4              0000000001      2  SCORE3\n",
       "...                   ...    ...     ...\n",
       "693655  ZeicheMasZeuchnis      4  SCORE1\n",
       "693656  ZeicheMasZeuchnis      6  SCORE5\n",
       "693657  ZeicheMasZeuchnis      5  SCORE7\n",
       "693658  ZeicheMasZeuchnis      1  SCORE3\n",
       "693659  ZeicheMasZeuchnis      2  SCORE8\n",
       "\n",
       "[693660 rows x 3 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gamescoringtiles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SCORE3    83727\n",
       "SCORE7    83435\n",
       "SCORE8    83322\n",
       "SCORE4    83270\n",
       "SCORE2    83226\n",
       "SCORE5    83188\n",
       "SCORE6    82770\n",
       "SCORE1    71728\n",
       "SCORE9    38994\n",
       "Name: tile, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gamescoringtiles['tile'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19\n",
      "option-errata-cultist-power                 0.106882\n",
      "option-strict-leech                         0.106427\n",
      "option-strict-darkling-sh                   0.103665\n",
      "option-strict-chaosmagician-sh              0.103642\n",
      "option-email-notify                         0.100557\n",
      "option-mini-expansion-1                     0.089982\n",
      "option-shipping-bonus                       0.083785\n",
      "option-variable-turn-order                  0.075660\n",
      "option-temple-scoring-tile                  0.053650\n",
      "option-fire-and-ice-factions/ice            0.039872\n",
      "option-fire-and-ice-final-scoring           0.039481\n",
      "option-fire-and-ice-factions/volcano        0.039336\n",
      "option-fire-and-ice-factions/variable_v5    0.024844\n",
      "option-maintain-player-order                0.021085\n",
      "option-fire-and-ice-factions/variable_v3    0.004580\n",
      "option-fire-and-ice-factions/variable       0.003220\n",
      "option-fire-and-ice-factions/variable_v4    0.002017\n",
      "option-fire-and-ice-factions/variable_v2    0.001303\n",
      "option-loose-adjust-resource                0.000012\n",
      "Name: option, dtype: float64\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>option</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>option-strict-darkling-sh</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>option-email-notify</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>option-shipping-bonus</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>option-strict-chaosmagician-sh</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>option-errata-cultist-power</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075569</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>option-email-notify</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075570</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>option-strict-chaosmagician-sh</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075571</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>option-errata-cultist-power</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075572</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>option-mini-expansion-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075573</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>option-strict-leech</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1075574 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      game                          option\n",
       "0               0000000001       option-strict-darkling-sh\n",
       "1               0000000001             option-email-notify\n",
       "2               0000000001           option-shipping-bonus\n",
       "3               0000000001  option-strict-chaosmagician-sh\n",
       "4               0000000001     option-errata-cultist-power\n",
       "...                    ...                             ...\n",
       "1075569  ZeicheMasZeuchnis             option-email-notify\n",
       "1075570  ZeicheMasZeuchnis  option-strict-chaosmagician-sh\n",
       "1075571  ZeicheMasZeuchnis     option-errata-cultist-power\n",
       "1075572  ZeicheMasZeuchnis         option-mini-expansion-1\n",
       "1075573  ZeicheMasZeuchnis             option-strict-leech\n",
       "\n",
       "[1075574 rows x 2 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(len(pd.unique(gameoptions['option'])))\n",
    "print(gameoptions['option'].value_counts() / len(gameoptions))\n",
    "gameoptions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "vp              12839907\n",
       "build:D          6570095\n",
       "leech:pw         5716489\n",
       "leech:count      5716489\n",
       "upgrade:TP       3543848\n",
       "                  ...   \n",
       "action:ACTH5        1852\n",
       "order:6              912\n",
       "order:7              636\n",
       "action:ACTH4         592\n",
       "action:ACTH3         314\n",
       "Name: event, Length: 91, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gameevents['event'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "126fe960806d587c78546b30f1a90853b1ada468    77158\n",
       "95a66999127893f5925a5f591d54f8bcb9a670e6    21052\n",
       "be8f6ebf549404d015547152d5f2a1906ae8dd90    16184\n",
       "fdb13a13cd48b7a3c3525f27e4628ff6905aa5b1     1210\n",
       "224736500d20520f195970eb0fd4c41df040c08c        3\n",
       "735b073fd7161268bb2796c1275abda92acd8b1a        2\n",
       "c07f36f9e050992d2daf6d44af2bc51dca719c46        1\n",
       "30b6ded823e53670624981abdb2c5b8568a44091        1\n",
       "b109f78907d2cbd5699ced16572be46043558e41        1\n",
       "Name: base_map, dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "games['base_map'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[126fe960806d587c78546b30f1a90853b1ada468](https://terra.snellman.net/map/126fe960806d587c78546b30f1a90853b1ada468)\n",
    "is the same as [91645cdb135773c2a7a50e5ca9cb18af54c664c4](https://terra.snellman.net/mapedit/91645cdb135773c2a7a50e5ca9cb18af54c664c4)  \n",
    "[95a66999127893f5925a5f591d54f8bcb9a670e6](https://terra.snellman.net/map/95a66999127893f5925a5f591d54f8bcb9a670e6) is different  \n",
    "[be8f6ebf549404d015547152d5f2a1906ae8dd90](https://terra.snellman.net/map/be8f6ebf549404d015547152d5f2a1906ae8dd90) is re-balanced.  \n",
    "\n",
    "All other maps aren't present here so we'll use those"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating the new datasets - some functions to help us make these later <a name=\"datasetCreate\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Victory Points (target) dataset <a name=\"VPData\"></a>\n",
    "Let's make some functions to help make the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def makenewdf():\n",
    "    \"\"\"make an empty dataframe, organised in the way we want the target data, ready to be populated\"\"\"\n",
    "    validfactions = ['witches', 'auren', 'swarmlings', 'mermaids', 'cultists', 'halflings', 'dwarves', 'engineers', 'chaosmagicians', 'giants', 'fakirs', 'nomads', 'darklings', 'alchemists']\n",
    "    dfcols = ['game'] + validfactions\n",
    "    vpdf = pd.DataFrame(columns=dfcols)\n",
    "    \n",
    "    return vpdf, dfcols, validfactions\n",
    "\n",
    "vpdf, dfcols, validfactions = makenewdf()\n",
    "\n",
    "sensecheck = False\n",
    "if sensecheck:\n",
    "    newdf = pd.DataFrame([[np.nan] * 15], columns=dfcols)\n",
    "    newdf['auren'].replace({np.nan: 'test'}, inplace=True)\n",
    "    vpdf = vpdf.append(newdf, ignore_index=True)\n",
    "    print(len(vpdf))\n",
    "    vpdf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_vp_from_game(singleGameEvents):\n",
    "    \"\"\"Input game events for a single game. This is a pd.DataFrame. \n",
    "    Output a row where each faction in the game has its vp populated (the rest are nans)\n",
    "    \"\"\"\n",
    "    newdf = pd.DataFrame([[np.nan] * 15], columns=dfcols)\n",
    "    \n",
    "    # assign the game number\n",
    "    gameno = list(pd.unique(singleGameEvents['game']))\n",
    "    \n",
    "    # assert len(gameno) == 1, 'More than 1 unique game was found'    \n",
    "    try:\n",
    "        newdf['game'].replace({np.nan: gameno[0]}, inplace=True)\n",
    "    except:\n",
    "        print(f'DEBUGGING: len of table is {len(singleGameEvents)}')\n",
    "        print(f'DEBUGGING: gamnos list: {gameno}')\n",
    "        print(singleGameEvents)\n",
    "        raise\n",
    "        \n",
    "    # find factions - there are some artifacts in the data. E.g. the \"faction\", \"all\". We need to filter them out.\n",
    "    rawfactions = list(pd.unique(singleGameEvents['faction']))\n",
    "    verifiedfactions = [rawfaction for rawfaction in rawfactions if rawfaction in validfactions]\n",
    "    \n",
    "    for faction in verifiedfactions:\n",
    "        vpfaction = sum(singleGameEvents[(singleGameEvents['event'] == 'vp') & (singleGameEvents['faction'] == faction)]['num'])\n",
    "        newdf[faction].replace({np.nan: vpfaction}, inplace=True)\n",
    "        \n",
    "    return newdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time elapsed: 3.538278818130493s\n",
      "Time to do all games at current speed: 113.62985853380627hrs\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>witches</th>\n",
       "      <th>auren</th>\n",
       "      <th>swarmlings</th>\n",
       "      <th>mermaids</th>\n",
       "      <th>cultists</th>\n",
       "      <th>halflings</th>\n",
       "      <th>dwarves</th>\n",
       "      <th>engineers</th>\n",
       "      <th>chaosmagicians</th>\n",
       "      <th>giants</th>\n",
       "      <th>fakirs</th>\n",
       "      <th>nomads</th>\n",
       "      <th>darklings</th>\n",
       "      <th>alchemists</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>62.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         game  witches  auren  swarmlings  mermaids  cultists  halflings  \\\n",
       "0  0000000001      NaN    NaN         NaN     118.0       NaN       62.0   \n",
       "\n",
       "   dwarves  engineers  chaosmagicians  giants  fakirs  nomads  darklings  \\\n",
       "0      NaN        NaN             NaN     NaN     NaN     NaN        NaN   \n",
       "\n",
       "   alchemists  \n",
       "0         NaN  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sense check\n",
    "start = time.time()\n",
    "singlegame = gameevents[gameevents['game'] == '0000000001']\n",
    "singlegamefactions = (list(pd.unique(singlegame['faction'])))\n",
    "vpforgame = get_vp_from_game(singlegame)\n",
    "end = time.time()\n",
    "\n",
    "print(f'Time elapsed: {end-start}s')\n",
    "print(f'Time to do all games at current speed: {((end-start)*len(pd.unique(gameevents[\"game\"]))/3600)}hrs')\n",
    "vpforgame"
   ]
  },
  {
   "attachments": {
    "a71ce2ba-dd59-49b1-aefc-8b6e1c14f8da.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That checks out (not inc the starting 20 vp)\n",
    "\n",
    "![image.png](attachment:a71ce2ba-dd59-49b1-aefc-8b6e1c14f8da.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features dataset <a name=\"featuresData\"></a>\n",
    "We'll want:\n",
    "1. Boolean of bonus tiles (in game or not)  <span style=\"color:green\">- DONE</span>\n",
    "2. One-hot of round tiles, for each round  <span style=\"color:green\">- DONE</span>\n",
    "3. Boolean of factions already picked (more info if any have been picked yet) <span style=\"color:orange\">- removed as would require 1 row for each faction. We want 1 line per game for now</span>\n",
    "4. One-hot for different map variants we'd like to use  <span style=\"color:green\">- DONE</span>\n",
    "5. One-hot player count (from 2, 3, 4 or 5 players)  <span style=\"color:green\">- DONE</span>"
   ]
  },
  {
   "attachments": {
    "942cd242-5b8b-432f-8843-51185679848e.png": {
     "image/png": "iVBORw0KGgoAAAANSUhEUgAAAIoAAAEKCAYAAADNQM0IAAAXN0lEQVR4Ae2d/ZHjKgzAU2NmtpUrYCc1vA6ypewVcn9sGX6DhNAHOMGbQIDoZt4ztkEI6WdBguI9bf7PLVBhgVNFHa/iFtgcFIegygIOSpWZvJKD4gxUWcBBqTKTV3JQnIEqCzgoVWbySg6KM1BlAQelykxeyUFxBqos4KBUmckrvTUo35+X7fsOAz9f5+30cd1+7tQb6vbfy3Y6xbH9u27n03m7/ntMw4ag/GzXj9N2Oon/Pu+55bHB5K2/t8vptF3+ijtguNN2+vzeABQw6mk7f5VRGB+UwhhnBIUdEMHpCkvBiMRMBCQAc+vflKDIAc0SURiUbduCc2IY//4MT/U1RR2qB46hKKRCPjo9Ragb9yiCQB8k6xSjxq8iymW7iOhI8tEfWi95z47xJKeAzIH4IHF7LZfsJhkI5eIY70SUXRuTbchm4iFqPvUQAGFQ8unEAeq5EwfA64ZQh9rLMhqI6qFBqd4Gg9X32PjazNVrFOlg6YQNnav75qnOjlGOH/WU49eg7I9Xj2HbcPxqjFJHA+S+jXX/WxjbJ6/NOoKiHQpGFMSCYh8MBpgjRSAcRHGakEaBRnLABSNaO985V86FukKmcUK4Lccly9A0jWfbboNyY7yZvkIfuidtonREuQnsUD/ppP1DoujYHJQ0VVDojz1nRoxPp6wP5TTF4EDofhosGEUsmGPYxCesYEQaeeUxB0WAqKIXCoT68QHIxpiccg+UIGtnvJnehTHeAYVsmI5kYxgP21JGqeagJIeaAWZGjKBI5UwTPpVPiTQ+14ilghGzOrcv5KAImVIP6hHWXrhAzsYodc3aCgCtSlldWUHoQ5fvgFJlYylj21pmuBXCHA3EhGi6DE6hz/9wkeZJOsaaynBoKAXk30v8SHxbB+r31tGCos9RfpoSQS+zRpHTqwTFri3AMdT21nittoUxSicrW8V1YtHG39vF6irqDRRR0AAIC4c/AgCeTlqN2+9GTJhOXzYFkdEBIcySLGvqW+dWHyUbGiKoFMbl03ozohjd6BMgtb89XqOxHeMNUEJLOya0S4Q+2VgutJtGFDMYP53aAg0jytR2ceWNBRwUYxA/LVvAQSnbxa8aCzgoxiB+WraAg1K2i181FnBQjEH8tGwBB6VsF79qLOCgGIP4adkCDkrZLn7VWMBBMQbx07IFuoHy8/Oz/ffff9ufP3/u/hfqhfr+bxwLdAMlOP/7+3Z+Kpkl1Av1/d84FugGSogkR/4drX9Ettc9bgEH5bjN3rKFgzKS22UeyUh6tc1w0yM9OpXU1bfJNiHhSSfcoBY6uSivQ/d1W0jwkVlfekjPP3NQNvikc8SyR0CRmWvgXEoWhg7LqZIhI42yyTCR+bxdPs/wC0LS00EhS3TMcKtzPCtWV/9Oviil/SlwsA8NVIApRBM6ijoiouj0RIo+COLlK/zGN6Zwqv7wPqVK2h9ygR7UDo70m6Sgg46Y8oFgS/UpTb5GyUHRUSC/n8wKScfkFAZEtpdl/v1LlJASuCMICQ7TZ6oX2mHd5HCYagg4yu8lnSIkCVTWMY2hY2EJUNLTSk9mMi4am6cYYVmVnS6dwGUJCpRFVjpLQufLPqBuAodrhpJMuJZlqCXXKEo/lBHqJ8i02OZnS4CijSefaFk2ttyJKKEWAUJHaqmmngRCDoqNPqpdgDmCnDk+A4V/jUAPgx4radb+2A2UNt/MlkGQDt57uvV1jiJocjy3i1t2h+w3B0XKBkgSVEcjCk1D3POrSt1AabPXIx1GJrTX0JHqSYQnN//UI182A84WT//P10W8jEb2YUHR5wCKmQopouBvjhgGjDx0jn2kumF4/67bZec9LjT6VsduoLQZQDQmrU3iUUEBHaPzKHyXv0cRi0rZhpwc4Uoy6Hr24zOzjoApjqaQ83YOr89IbfVrKy6f4k1JoIMdn9WxjVVLUicHpTSk3td0BOnde6/+HJSHLe2gPGxCF7COBTyirOPLpiNxUJqadx3hDso6vmw6EgelqXnXEe6grOPLpiPpBkqbb2ab2saFCwt0A6XNXo8YiRebWqAbKHWJSDzWo/W5pZdaWMBBaWHVBWU6KI2dqnePG3fWUPzkoNjd1fGy8BUohay1hr59quglQJFpBTJpCC2Fm3ayDr17ltMXQ502WfgOykFejy5O6+pjRMkhoOQf/Rc9pMoaKATlXhY+JUdTTgr2m+8eBzhIJwYlj34MqtRszPKaESUlBhVAIj/s5MwCQLG9LBMkBAD8lOIr/Oi+FpTSH0ogZcY/LgEKPeHpaEApPrlqvUARJTiMywoUmfis/OqgKHM8elI3lXAvdfVLEUNek2WWDaWdiBLuESB0hPoBFJEkzdIcFLbFE0ptvpktgyAdDOWCg/V1jiI4VDxXWfgeUZ5AQYWINns9JVDsNXzieW1Bv8g7loVPaxGWQ3+2xPQHkaq0mA1GyqNPhemGqNJtjdJmtOiktDZpmYUfBhAhoP4SNPL6x3W7Fj/1oAUgkkU9i2unNoZ6WOrkoDw8fhdQaQEHpdJQ717NQXl3AirH76BUGurdqzko705A5fgdlEpDvXs1B+XdCagcv4NSaah3r9YNlDbfzL67+/qNvxsobfZ6+hnq3XvqBkrdbjC742h9bumlFhZwUFpYdUGZDsqCTm0xpMlBKe0el95zhtv7tOtbfodbnsEv81qS8dW73Dg3N91ftLAEKGm7n7LTVKJSfT6KSlQiWSmtkvJY3gcOyfxyoNhXcupMNh66vh5gCpGIjlhPRxR9jyVxCeqnN1SuBdRyoGjnmuwz9mlMQiJnMgSyvSzbpCWYxkS0QUhIXog+V/FeWtnxnOUlQOG1R3yfa3IgglLMJIOsNFrPMCg3s/B3p7QbQM7JRab1EqDINQr96RK8dsOBAApFAAlKfRY+RBGA5waQmcnnvNANlDbfzJZBkFMGO1M7SF/XoFBUUYtbBRbK4n7Keuge5z7rBkqbvZ6Sg+y1AAFnxYO74kdcnpIsKDGqhIWpmcb4HOWSDIBG/JkW/e78uSEJ2ncDpY2pEAq7RtFTUegZncr1aG1CWuWgpDYJlFyO7Qd+Z+yfesiofnxHC0weUd7RZa8Zs4PyGrtP16uDMp3LXqOwg/Iau0/Xq4Myncteo7CD8hq7T9ergzKdy16jcDdQ2nwz+xqjvWOv3UBps9fzji57zZi7gXI0q/5o/deY7316dVDex9cPjdRBech8obHeRbbi+IXE9s5c55ODUto9tjvD7Mzbu8eVWfiZfx2UzCSPXDi65qirb3NPSq80R0eqlICdfBSVqFTKwi8a4DYoxSYTXlwiouQQUIpjCRz0UjnDLTidIxJnsGEbnW9C9SIoX9ftTLkoIrdWTT2Up/t12VJ0U/ku4xK0HCjauXnESa5QqY0MiGwvy/AzEAFAOMfsNgSF32qt+8xBkVlzEbK/SathC0uAkp5OeqLTU4pOo3RF5QV6uv+FqwyKLEtQoCxSHVlW7uwAB0W5HBSKRChB3Wehw5WWAIWcgtaVT7QsG9vvRJRQiwChI7UEpxKMKbo8Bortg/oa7dgNlDbfzJZBkMaHcnIqm19flxEl1MFzu7gVrbfrB0WNx0DxiMJWhVKbvZ4SKPYaOlJFnZ1PPVeYhlBxAClEjziN6ax62ccDoKjpzxhssNNuEaXNuNFhdo2ioICO0ZlcT68TKIJIUPAag0J/Xi7JSOugo6DEXzPGKay4fmpjrIekTg7KQ2Pv33iiCGKN46BYi7Q8d1BaWtdlj2ABjygjeGECHRyUCZw0gooOyghemEAHB2UCJ42gooMyghcm0KEbKG2+mZ3Awouo2A2UNns9i3hhgmF0A6UuY40tdrQ+t/RSCws4KC2suqBMB2VBp7YY0uSglHaP7c5wMFvN7vH9LHyVuHQ6xVTIFm4J+pbG0aKvOplLgCLTCnRCEkMi61DKAG/xo2NsohLIkukEMgGqyQafBN9BqUK4bjErE4iiWEhKapOFrxXXeSiYqXaFzDfMWWFHa+BQ5wQpAMf6Yh+liGKuhXEKcEP/6mHQyj58tmZESVGgABKZTDmInSCdKsvULB0BSIYBpyU+h7bkSOlU6NckRFG9JJz1SZc2DZjuT9/jNs8rLQFKyjqjxGcDSnp6pd3U1CEdw+UiKORos0bBiPLNPezIDzLPX9ftEjP68fyH20GJdZA3WJ9w/7JdQVZoi+eid9nsKeUlQNEhV0YRWTb2AkdSyNeOIYfQ0bSMp6ENL2gzUNRPQPiJ//7EqBPqB4DpqPvQ+qR7FJmC7uFhCMcQjeiYKj6/0A2UNt/MlkGQDoZyFtrtLwitY/DcLm6t+YOTCdIMFABRT0UQSUiX4PTPS4osWrbVh+6G6xhJMEqG8Z+3oCfpQTWffewGSpu9nhIo9lowLjsUDAjrC44GpeRqAExk4cPTm6a0Lf69H5ZhQbHn9EmLHYp6yQUpO3cPFBzbSfwQjfQsTq8s8OFSN1Ae1rQogAynM9vZGdQoOoXWMNl3FCXHxDYJjrwv6RwAI8k/qU8kqEXeR2ijdc37sCDJaAlyIXLRFErjff5xclCeb5DfSswiyG8FDdrOQXmSYxyUJxnSxcxtAY8oc/uvm/YOSjdTz92RgzK3/7pp76B0M/XcHTkoc/uvm/bdQGnzzWw3O719R91AabPX8/b+62aAbqDUJSLxuI/W55ZeamEBB6WFVReU6aD82qm4aSg3Bn8taoKGk4NS2G3NdoaDF+JOcNrd5TwR9BHd19eznVrlUAdFmeNZJ0fXHHX1ERS5VQ/OpeQgUB4dKutQbghHg1AHE4DoLZChqYPC3l8iouQQcH5GDg4OXl9HUK7/6CjqpHwUTlaSOboAm8lmg9aQHIV64M5yOUOfXTF2aTlQdBTII05yh0r4YUBke1mm6YuhDG10hhvfw1xYOgdQxJSoIU0aDV1YAhT5hEM5RQEEhacY4QsVBRgUmRapQBERAqVoUGA6S1OelIfQyCkN0ioFOEKrYYtLgEJPLlpZRhFZNj7YiSihFgFCR2gZQEkghCsGFJl1b+rmSU0aJKPZkKfdQGnzzWwZBOlgKCsHox/0des4PFdZ+PciSkAn5sCGo4xiGSgqmg3JRaZUN1Da7PWUQLHX8MlXUQecLp1pQYlRRWbh2wiSyYiL3Y/zdhZZ8sHiFhR7nnllwAvdQGkzdoTCrlEUFNAxwsL19Pclcl3CesY2ab2zbfSxGtdB4Tc5ErbQMuoj2xAo6TucUoY+9zpqaXJQRjWr1mvGCKJHsG0OirVIg3MHpYFRXeSYFvCIMqZfhtPKQRnOJWMq5KCM6ZfhtHJQhnPJmAo5KGP6ZTituoHS5pvZ4ey5rELdQGmz17OsX4YbWDdQ6jLW2D5H63NLL7WwgIPSwqoLynRQFnRqiyFNDkpp99juDAez1ewe33kX/oQ5JM8EZglQZFqBTkhiSGQdShfg5CLMR1GJSiLTDQzuoDyTu31ZRxendfVtkhLljDw7Cz+PXAEygvL6yW+lJPjyHWOpK0a4y9d1O1OeSiELb9+a/e+sGVFS4pB0jjHuTs4sOD+2l+VSQjTcl8lLMl1SyaefehDAcSpMcNzQ06j9qtMlQOHMtfhkG1DoKVdGVlMJTj3Xf6EGl6tASc6mtpT1hs6nvpUsm1ZJ05ySpbR9+ckSoKj1R0xHxGs3nlT1xDMcwSPkVDqClxRY6De4r5y7B4e+TotrggikhWikZGEfo/y/GyhtvpktgyAdnDuz5GQNCkUVtbitAiXIoYgippvQVkFg6hGcqs4oiKAe3UBps9dTAsVeQ6eoqJNl0FtQKrLwC84tQRkWteeQmf8l/9SKBcWejwVJ0KYbKG2GjlDYNYp2SugZHcH17HctOSipTVrvCHhi1AAw6FMLHGmxKkYLUJb6409KQa9cZyFjgOLkoLzWgqUIUqfR+BHEjsNBsRY5cO6gHDCWV30PC3hEeQ8/PzxKB+VhE76HAAflPfz88CgdlIdN+B4CHJT38PPDo+wGSptvZh8evwuotEA3UNrs9VSO0qs9bIFuoNQlIvF4jtbnll5qYQEHpYVVF5TpoCzo1BZDmhyU0u6x3akNZqvZPb6Thd/C+hPJXAIUuUWfb9QhJLLOr7LwJ3JqC1WXAwUh4LyQHBw0o75O+Sh0FHVEPgpk1qf8E4xcWg624wx8hHSmbPs9yJYDBRyXnGuz3YQZdnJmZXtZBgBlquLfS3zpsIaL0ig5UVu+LvSGPkK1EYtLgMKZa+2y8AEa86JhciikO1Kqo0qSjhHlL9Xk3wLJxEi+O25pCVDU+qNVFn5YEosfeqmMeQGHgqbws4wsMo3LhtKsGyhtvpkth3I5ZUBZThlx+Pp6efpQWfjKbLZfah+OvD6iT1vyZxm6XyV06JNuoLTZ67EOC7a21zD8q6jziyz8n6/LhuuOUh84pYRse/XnVrKIgrpIcIamQyjXDRTR5xOLCIVdoygooDd0ENez37WE+6VrJ3Z8hCvJSAvmOBxYHIvf9BT7HT/bfs85k4OyN6xRrs8bQawFHRRrkaeeOyhPNacLG98CHlHG99EQGjooQ7hhfCUclPF9NISGDsoQbhhfCQdlfB8NoWE3UNp8MzuEDd9CiW6gtNnreQsfDTHIbqAczao/Wn8Iay6shIOysHOfOTQH5ZnWXFjW5KCUdo/tLnDwXs3ucX0Wvk5OWpgOMbQlQJFpBXliEEIi6/w2Cx9kx+RqJU8YdNXicqC0zMInCEoRBdIkP6/b9YPe9liKbCRhvuNyoMBTn5KKbLabcNDRLHzRdBcUkfyURzYhYMLiEqCkrDP6zY0BpZh6CKDQUx+mp7ysoWPv7oKS+qW3VpNMbjtraQlQ9HpBRhFZNi7aiSihFgFCR9MSsvF1nzFDX4ICC2gHxdru7nmbb2bLIEgHQ/nJWfgeUe66+/cV2uz1lECx1+o/9XCWvXiduYoSOP4aUEIdnZH/e9uN0LLb1NNmsAiFXaPYaaHuexQ7TcTvXgQoEJ1oHQRHbgNgyHuFKNbGBn2kTg5KHyPV9LJaBLFjdlCsRX557qD80nDebC0LeERZy5/NRuOgNDPtWoIdlLX82Ww0Dkoz064l2EFZy5/NRtMNlDbfzDaziws2FugGSpu9HjMaP21mgW6gHM2qP1q/mYVcMFjAQXEQqizgoFSZ6VYl3DwsJkfRmyTFxuItSSPfmxyU0u4x7+iy4eNOcNrdtXXovr4u81pYli05KNYiD50fXXPU1be5J6UX/qIjVerBzlsh7etCnwHKQ0YbqPESESWHgN/1eizDLUDFUcWConNOqF6MKF/X7UwRS+SiqF1lytP9umwph4amJbr3T9ABQONYUM7rsvyXA0U7N484yQ07ObOyvSxnb5xW78Lff999DorMfNPTVqgroZfnIMdCLIBM42pUWAKU9HTSE01PaXw5cXGhqZ5gGUm4LEGBsnorNXlEOztczRxM+qg+sT0AQPdDBEnOZz1IpkqtLMgijVoclwBFPoX6zdXHI0owMgFCRzI8PtXxB17KofpFxEdA0X0IOBQ0r8/y7wZKm29myyBI40M5OZVcbhe9wkFQBc/t4la0hl8EIqCh7u9BURFFRKNwXUZCW29bNaK02espgWKvoSNV1Nn51HMrC3//XfgPgFJydrj2cd7OZpqzoNhzhrhNqVtEaaM+QmHXKAoK6BidyfXoEwtpFe6XromFZ4QryaB1RfaHEe6tUei3yXiUUQO1iWNK8vEqgEFrsHAsREkaTYvj5KC0MElDmaUIUtld7whi1XJQrEVanjsoLa3rskewgEeUEbwwgQ4OygROGkFFB2UEL0ygg4MygZNGUNFBGcELE+jgoEzgpBFUdFBG8MIEOjgoEzhpBBUdlBG8MIEODsoEThpBRQdlBC9MoIODMoGTRlDRQRnBCxPo4KBM4KQRVHRQRvDCBDo4KBM4aQQVHZQRvDCBDg7KBE4aQcX/AQ2Rq5P5RHWPAAAAAElFTkSuQmCC"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These are the bonus tiles available:  \n",
    "![image.png](attachment:942cd242-5b8b-432f-8843-51185679848e.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['pass:BON10', 'build:D', 'pass:BON3'], dtype=object)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.unique(gameevents[(gameevents['round'] == 0) & (gameevents['game'] == 'ytuwertqwtr')]['event'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But it looks like we only have details on the bonus tiles selected in the middle of the game. Where can we get information about all bonus tiles in game!? I'm not sure, so I'll use the only info that's available - which is the bonus tiles that players select. However, if there's a bonus tile that was never selected during a game, it won't be shown. If we go onto the Snellman website and re-watch an old game, we can see what bonus tiles are available in the GUI, so this info must be stored somewhere (as game history GUI must be using the same data we're looking at...?)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time elapsed: 3.645721197128296s\n",
      "['BON8', 'BON6', 'BON3', 'BON10', 'BON9']\n"
     ]
    }
   ],
   "source": [
    "# working out the code to get the bonus tiles & timing this\n",
    "start = time.time()\n",
    "allevents = list(pd.unique(gameevents[gameevents['game'] == 'ytuwertqwtr']['event']))\n",
    "bonustiles = [event[5:] for event in allevents if event.startswith('pass:BON')]\n",
    "end = time.time()\n",
    "print(f'Time elapsed: {end-start}s')\n",
    "print(bonustiles)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['game', 'r1_SCORE1', 'r1_SCORE2', 'r1_SCORE3', 'r1_SCORE4', 'r1_SCORE5', 'r1_SCORE6', 'r1_SCORE7', 'r1_SCORE8', 'r1_SCORE9', 'r2_SCORE1', 'r2_SCORE2', 'r2_SCORE3', 'r2_SCORE4', 'r2_SCORE5', 'r2_SCORE6', 'r2_SCORE7', 'r2_SCORE8', 'r2_SCORE9', 'r3_SCORE1', 'r3_SCORE2', 'r3_SCORE3', 'r3_SCORE4', 'r3_SCORE5', 'r3_SCORE6', 'r3_SCORE7', 'r3_SCORE8', 'r3_SCORE9', 'r4_SCORE1', 'r4_SCORE2', 'r4_SCORE3', 'r4_SCORE4', 'r4_SCORE5', 'r4_SCORE6', 'r4_SCORE7', 'r4_SCORE8', 'r4_SCORE9', 'r5_SCORE1', 'r5_SCORE2', 'r5_SCORE3', 'r5_SCORE4', 'r5_SCORE5', 'r5_SCORE6', 'r5_SCORE7', 'r5_SCORE8', 'r5_SCORE9', 'r6_SCORE1', 'r6_SCORE2', 'r6_SCORE3', 'r6_SCORE4', 'r6_SCORE5', 'r6_SCORE6', 'r6_SCORE7', 'r6_SCORE8', 'r6_SCORE9', 'BON1', 'BON2', 'BON3', 'BON4', 'BON5', 'BON6', 'BON7', 'BON8', 'BON9', 'BON10', '2players', '3players', '4players', '5players', 'map1', 'map2', 'map3']\n",
      "72\n"
     ]
    }
   ],
   "source": [
    "def emptyfeaturesdf():\n",
    "    \"\"\"make an empty dataframe, organised in the way we want the feature data, ready to be populated\"\"\"\n",
    "    colnames = ['game']\n",
    "    uniqueScoreTiles = np.sort(pd.unique(gamescoringtiles['tile']))\n",
    "    \n",
    "    # One-hot of round tiles, for each round\n",
    "    for gameround in range(1, 7):\n",
    "        roundstr = f'r{gameround}'\n",
    "        for tile in uniqueScoreTiles:\n",
    "            colnames.append(roundstr + '_' + tile)\n",
    "    \n",
    "    # Boolean of bonus tiles\n",
    "    for bon in range(1, 11):\n",
    "        colnames.append(f'BON{bon}')\n",
    "        \n",
    "    # One-hot player count (from 2, 3, 4 or 5 players)\n",
    "    for player in range(2, 6):\n",
    "        colnames.append(f'{player}players')\n",
    "    \n",
    "    # one hot of the map used\n",
    "    \"\"\"126fe960806d587c78546b30f1a90853b1ada468 - map1\n",
    "       95a66999127893f5925a5f591d54f8bcb9a670e6 - map2\n",
    "       be8f6ebf549404d015547152d5f2a1906ae8dd90 - map3\n",
    "    \"\"\"\n",
    "    colnames = colnames + ['map1', 'map2', 'map3']\n",
    "    \n",
    "    featuresdf = pd.DataFrame(columns=colnames)\n",
    "    \n",
    "    return featuresdf, colnames\n",
    "\n",
    "featuresdf, featcolnames = emptyfeaturesdf()\n",
    "print(featcolnames)\n",
    "print(len(featcolnames))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_features_from_game(singlegameevents, singlegamemeta, singlegameST, singleendplayers=None):\n",
    "    \"\"\"\n",
    "    Inputs:\n",
    "        singlegameevents <pd.DataFrame>   - is game events for a single game \n",
    "        singlegamemeta   <pd.DataFrame>   - is a single row from `games` that gives map & player count\n",
    "        singlegameST     <pd.DataFrame>   - is a single row from `gamescoringtiles` that gives... score tile (suprisingly)\n",
    "        singleendplayers <pd.DataFrame>   - is a single row from `end players` that gives the amount of players at end of game, after dropouts\n",
    "    Return:          <pd.DataFrame>   - a row where features have been found (will be sparse)\n",
    "    \"\"\"\n",
    "    newdf = pd.DataFrame([[0] * len(featcolnames)], columns=featcolnames)\n",
    "    \n",
    "    # assign game string\n",
    "    singlegamemeta.iloc[0]['game']\n",
    "    newdf['game'].replace({0: singlegamemeta.iloc[0]['game']}, inplace=True)\n",
    "    \n",
    "    # find the round tiles for each round\n",
    "    for gameround in range(1, 7):\n",
    "        roundstr = f'r{gameround}'\n",
    "        scoretile = roundstr + '_' + singlegameST[singlegameST['round'] == gameround]['tile'].values[0]\n",
    "        newdf[scoretile].replace({0: 1}, inplace=True)\n",
    "        \n",
    "    # Boolean of bonus tiles\n",
    "    uniqueevents = list(pd.unique(singlegameevents['event']))\n",
    "    bonustiles = [event[5:] for event in uniqueevents if event.startswith('pass:BON')]\n",
    "    for bontile in bonustiles:\n",
    "        newdf[bontile].replace({0: 1}, inplace=True)\n",
    "       \n",
    "    # One-hot player count (from 2, 3, 4 or 5 players)\n",
    "    if singleendplayers is None:\n",
    "        noplayers = singlegamemeta.iloc[0]['player_count']\n",
    "        print('gamemeta used for player count')\n",
    "    else:\n",
    "        noplayers = singleendplayers.iloc[0]['endplayers']\n",
    "        \n",
    "    players = f'{noplayers}players'\n",
    "    newdf[players].replace({0: 1}, inplace=True)\n",
    "          \n",
    "    # one hot of the map used\n",
    "    mapdict = {'126fe960806d587c78546b30f1a90853b1ada468': 'map1',\n",
    "               '95a66999127893f5925a5f591d54f8bcb9a670e6': 'map2',\n",
    "               'be8f6ebf549404d015547152d5f2a1906ae8dd90': 'map3'\n",
    "              }\n",
    "    basemap = singlegamemeta.iloc[0]['base_map']\n",
    "    gamemap = mapdict[basemap]\n",
    "    newdf[gamemap].replace({0: 1}, inplace=True)\n",
    "        \n",
    "    return newdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gamemeta used for player count\n",
      "Time elapsed: 3.5182363986968994s\n",
      "Time to do all games at current speed: 112.98620736837387hrs\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>r1_SCORE1</th>\n",
       "      <th>r1_SCORE2</th>\n",
       "      <th>r1_SCORE3</th>\n",
       "      <th>r1_SCORE4</th>\n",
       "      <th>r1_SCORE5</th>\n",
       "      <th>r1_SCORE6</th>\n",
       "      <th>r1_SCORE7</th>\n",
       "      <th>r1_SCORE8</th>\n",
       "      <th>r1_SCORE9</th>\n",
       "      <th>...</th>\n",
       "      <th>BON8</th>\n",
       "      <th>BON9</th>\n",
       "      <th>BON10</th>\n",
       "      <th>2players</th>\n",
       "      <th>3players</th>\n",
       "      <th>4players</th>\n",
       "      <th>5players</th>\n",
       "      <th>map1</th>\n",
       "      <th>map2</th>\n",
       "      <th>map3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 72 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         game  r1_SCORE1  r1_SCORE2  r1_SCORE3  r1_SCORE4  r1_SCORE5  \\\n",
       "0  0000000001          0          0          0          0          0   \n",
       "\n",
       "   r1_SCORE6  r1_SCORE7  r1_SCORE8  r1_SCORE9  ...  BON8  BON9  BON10  \\\n",
       "0          0          0          1          0  ...     0     1      1   \n",
       "\n",
       "   2players  3players  4players  5players  map1  map2  map3  \n",
       "0         1         0         0         0     1     0     0  \n",
       "\n",
       "[1 rows x 72 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sense check\n",
    "start = time.time()\n",
    "singlegame = gameevents[gameevents['game'] == '0000000001']\n",
    "singlegamemeta = games[games['game'] == '0000000001']\n",
    "singlegameST = gamescoringtiles[gamescoringtiles['game'] == '0000000001']\n",
    "vpforgame = get_vp_from_game(singlegame)\n",
    "featsforgame = get_features_from_game(singlegame, singlegamemeta, singlegameST)\n",
    "end = time.time()\n",
    "\n",
    "print(f'Time elapsed: {end-start}s')\n",
    "print(f'Time to do all games at current speed: {((end-start)*len(pd.unique(gameevents[\"game\"]))/3600)}hrs')\n",
    "featsforgame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Speeding up dataset creation <a name=\"speedUp\"></a>\n",
    "350 hours: that's way too long. How long will it take to split the table into 1500, 100 game tables and work on those?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time elapsed to make small table: 1.233781099319458s\n",
      "gamemeta used for player count\n",
      "Time elapsed to get data for single game now: 0.03003406524658203s\n",
      "Time elapsed to do all small tables: 0.39622194570700325hrs\n",
      "Time elapsed to do all data creation: 0.9645273198021783hrs\n",
      "Time elapsed to do all processing: 1.3607492655091815hrs\n"
     ]
    }
   ],
   "source": [
    "first100games = gameslist[:100]\n",
    "\n",
    "start = time.time()\n",
    "gameevents100 = gameevents[gameevents['game'].isin(first100games)]\n",
    "gamemeta100 = games[games['game'].isin(first100games)]\n",
    "gameST100 = gamescoringtiles[gamescoringtiles['game'].isin(first100games)]\n",
    "end = time.time()\n",
    "smalltablestime = end-start\n",
    "\n",
    "print(f'Time elapsed to make small table: {smalltablestime}s')\n",
    "\n",
    "start = time.time()\n",
    "singlegame = gameevents100[gameevents100['game'] == '0000000001']\n",
    "singlegamemeta = gamemeta100[gamemeta100['game'] == '0000000001']\n",
    "singlegameST = gameST100[gameST100['game'] == '0000000001']\n",
    "vpforgame = get_vp_from_game(singlegame)\n",
    "featsforgame = get_features_from_game(singlegame, singlegamemeta, singlegameST)\n",
    "end = time.time()\n",
    "datacreatetime = end-start\n",
    "\n",
    "print(f'Time elapsed to get data for single game now: {end-start}s')\n",
    "totalsmalltablestime = smalltablestime * len(pd.unique(gameevents[\"game\"]))/(100 * 3600)  # divide by 100 as that's the size of table, divide by 3600 for s -> hrs\n",
    "totaldatacreatetime = datacreatetime * len(pd.unique(gameevents[\"game\"]))/3600\n",
    "print(f'Time elapsed to do all small tables: {totalsmalltablestime}hrs')\n",
    "print(f'Time elapsed to do all data creation: {totaldatacreatetime}hrs')\n",
    "print(f'Time elapsed to do all processing: {totalsmalltablestime + totaldatacreatetime}hrs')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Damn. I've just saved ~200 times the amount of time by doing that. As a complete aside, for pure interest, I wonder what the best trade-off between spitting into smaller & smaller tables (which increases time to do this) is vs saving time to then filter with smaller tables, is? (will check this out later)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Progressed to 100th game\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "Progressed to 200th game\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "gamemeta used for player count\n",
      "no of unique games in table is: 200\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>witches</th>\n",
       "      <th>auren</th>\n",
       "      <th>swarmlings</th>\n",
       "      <th>mermaids</th>\n",
       "      <th>cultists</th>\n",
       "      <th>halflings</th>\n",
       "      <th>dwarves</th>\n",
       "      <th>engineers</th>\n",
       "      <th>chaosmagicians</th>\n",
       "      <th>giants</th>\n",
       "      <th>fakirs</th>\n",
       "      <th>nomads</th>\n",
       "      <th>darklings</th>\n",
       "      <th>alchemists</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>62.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>050615</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>122.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>124.0</td>\n",
       "      <td>101.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>051501</td>\n",
       "      <td>NaN</td>\n",
       "      <td>84.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>77.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>72.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>54.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>060303</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>108.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>121.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>78.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>128.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\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",
       "      <th>195</th>\n",
       "      <td>4pLeague_S7_D6L17_G7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>130.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>4pLeague_S7_D6L18_G2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>105.0</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>108.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>128.0</td>\n",
       "      <td>53.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>4pLeague_S7_D6L18_G5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>97.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>82.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>110.0</td>\n",
       "      <td>117.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>4pLeague_S7_D6L18_G6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>78.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>103.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>137.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>61.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>4pLeague_S7_D6L18_G7</td>\n",
       "      <td>95.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>122.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>90.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>120.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>200 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     game  witches  auren  swarmlings  mermaids  cultists  \\\n",
       "0              0000000001      NaN    NaN         NaN     118.0       NaN   \n",
       "1                  050615      NaN    NaN         NaN       NaN       NaN   \n",
       "2                    0512      NaN    NaN         NaN       NaN       NaN   \n",
       "3                  051501      NaN   84.0         NaN       NaN       NaN   \n",
       "4                  060303      NaN    NaN         NaN     108.0       NaN   \n",
       "..                    ...      ...    ...         ...       ...       ...   \n",
       "195  4pLeague_S7_D6L17_G7      NaN    NaN        48.0       NaN       NaN   \n",
       "196  4pLeague_S7_D6L18_G2      NaN  105.0         NaN       NaN       NaN   \n",
       "197  4pLeague_S7_D6L18_G5      NaN    NaN        97.0       NaN       NaN   \n",
       "198  4pLeague_S7_D6L18_G6      NaN    NaN        78.0       NaN       NaN   \n",
       "199  4pLeague_S7_D6L18_G7     95.0    NaN         NaN       NaN       NaN   \n",
       "\n",
       "     halflings  dwarves  engineers  chaosmagicians  giants  fakirs  nomads  \\\n",
       "0         62.0      NaN        NaN             NaN     NaN     NaN     NaN   \n",
       "1          NaN      NaN        NaN             NaN     NaN     NaN     NaN   \n",
       "2          NaN      NaN      124.0           101.0     NaN     NaN     NaN   \n",
       "3         77.0      NaN        NaN            72.0     NaN     NaN    54.0   \n",
       "4          NaN      NaN      121.0             NaN     NaN    78.0     NaN   \n",
       "..         ...      ...        ...             ...     ...     ...     ...   \n",
       "195        NaN      NaN        NaN            73.0     NaN     NaN   130.0   \n",
       "196        NaN      NaN        NaN           108.0     NaN     NaN   128.0   \n",
       "197        NaN     82.0        NaN             NaN     NaN     NaN   110.0   \n",
       "198      103.0      NaN        NaN           137.0     NaN     NaN     NaN   \n",
       "199      122.0      NaN        NaN            90.0     NaN     NaN     NaN   \n",
       "\n",
       "     darklings  alchemists  \n",
       "0          NaN         NaN  \n",
       "1        122.0         NaN  \n",
       "2          NaN         NaN  \n",
       "3         78.0         NaN  \n",
       "4        128.0         NaN  \n",
       "..         ...         ...  \n",
       "195       87.0         NaN  \n",
       "196       53.0         NaN  \n",
       "197      117.0         NaN  \n",
       "198       61.0         NaN  \n",
       "199      120.0         NaN  \n",
       "\n",
       "[200 rows x 15 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sense check for 100 games\n",
    "vpdf, _, _ = makenewdf()\n",
    "featdf, _ = emptyfeaturesdf()\n",
    "gamelengthlen = len(gameslist)\n",
    "gamesroundup = math.ceil(gamelengthlen / 100.0) * 100\n",
    "jj = 0\n",
    "\n",
    "for ii in range(100, 201, 100):     \n",
    "    ii = min(ii, gamelengthlen)    \n",
    "    print(f'Progressed to {ii}th game')\n",
    "    \n",
    "    next100games = gameslist[jj:ii]\n",
    "    jj = ii  # so that we don't get any repetitions at the very end, where our set will be smaller\n",
    "    \n",
    "    gameevents100 = gameevents[gameevents['game'].isin(next100games)]\n",
    "    gamemeta100 = games[games['game'].isin(next100games)]\n",
    "    gameST100 = gamescoringtiles[gamescoringtiles['game'].isin(next100games)]\n",
    "    \n",
    "    for game in next100games:\n",
    "        singlegame = gameevents100[gameevents100['game'] == game]\n",
    "        singlegamemeta = gamemeta100[gamemeta100['game'] == game]\n",
    "        singlegameST = gameST100[gameST100['game'] == game]\n",
    "        \n",
    "        if not len(singlegame) == 0:        \n",
    "            vpforgame = get_vp_from_game(singlegame)\n",
    "            featsforgame = get_features_from_game(singlegame, singlegamemeta, singlegameST)\n",
    "            \n",
    "            vpdf = vpdf.append(vpforgame, ignore_index=True)\n",
    "            featdf = featdf.append(featsforgame, ignore_index=True)\n",
    "    \n",
    "print(f\"no of unique games in table is: {len(list(pd.unique(vpdf['game'])))}\")\n",
    "vpdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "tags": []
   },
   "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>game</th>\n",
       "      <th>r1_SCORE1</th>\n",
       "      <th>r1_SCORE2</th>\n",
       "      <th>r1_SCORE3</th>\n",
       "      <th>r1_SCORE4</th>\n",
       "      <th>r1_SCORE5</th>\n",
       "      <th>r1_SCORE6</th>\n",
       "      <th>r1_SCORE7</th>\n",
       "      <th>r1_SCORE8</th>\n",
       "      <th>r1_SCORE9</th>\n",
       "      <th>...</th>\n",
       "      <th>BON8</th>\n",
       "      <th>BON9</th>\n",
       "      <th>BON10</th>\n",
       "      <th>2players</th>\n",
       "      <th>3players</th>\n",
       "      <th>4players</th>\n",
       "      <th>5players</th>\n",
       "      <th>map1</th>\n",
       "      <th>map2</th>\n",
       "      <th>map3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>050615</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0512</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>051501</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>060303</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>4pLeague_S7_D6L17_G7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>4pLeague_S7_D6L18_G2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>4pLeague_S7_D6L18_G5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>4pLeague_S7_D6L18_G6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>4pLeague_S7_D6L18_G7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>200 rows × 72 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     game r1_SCORE1 r1_SCORE2 r1_SCORE3 r1_SCORE4 r1_SCORE5  \\\n",
       "0              0000000001         0         0         0         0         0   \n",
       "1                  050615         0         0         0         0         1   \n",
       "2                    0512         0         0         0         0         1   \n",
       "3                  051501         0         0         0         0         0   \n",
       "4                  060303         0         1         0         0         0   \n",
       "..                    ...       ...       ...       ...       ...       ...   \n",
       "195  4pLeague_S7_D6L17_G7         0         0         0         0         0   \n",
       "196  4pLeague_S7_D6L18_G2         0         0         0         0         0   \n",
       "197  4pLeague_S7_D6L18_G5         0         0         0         0         0   \n",
       "198  4pLeague_S7_D6L18_G6         0         0         0         0         0   \n",
       "199  4pLeague_S7_D6L18_G7         0         0         0         0         0   \n",
       "\n",
       "    r1_SCORE6 r1_SCORE7 r1_SCORE8 r1_SCORE9  ... BON8 BON9 BON10 2players  \\\n",
       "0           0         0         1         0  ...    0    1     1        1   \n",
       "1           0         0         0         0  ...    1    1     0        0   \n",
       "2           0         0         0         0  ...    1    1     0        0   \n",
       "3           0         0         1         0  ...    1    0     0        0   \n",
       "4           0         0         0         0  ...    1    1     1        0   \n",
       "..        ...       ...       ...       ...  ...  ...  ...   ...      ...   \n",
       "195         0         1         0         0  ...    1    1     1        0   \n",
       "196         0         1         0         0  ...    1    0     1        0   \n",
       "197         0         0         1         0  ...    1    1     0        0   \n",
       "198         0         0         1         0  ...    1    0     1        0   \n",
       "199         0         0         1         0  ...    1    1     1        0   \n",
       "\n",
       "    3players 4players 5players map1 map2 map3  \n",
       "0          0        0        0    1    0    0  \n",
       "1          0        1        0    0    0    1  \n",
       "2          0        1        0    1    0    0  \n",
       "3          0        0        1    1    0    0  \n",
       "4          0        1        0    0    1    0  \n",
       "..       ...      ...      ...  ...  ...  ...  \n",
       "195        0        1        0    1    0    0  \n",
       "196        0        1        0    1    0    0  \n",
       "197        0        1        0    1    0    0  \n",
       "198        0        1        0    1    0    0  \n",
       "199        0        1        0    1    0    0  \n",
       "\n",
       "[200 rows x 72 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "featdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks good! We can soon do it with the full set of data... after we've done some cleaning.\n",
    "\n",
    "## Cleaning and filtering the data <a name=\"cleanUp\"></a>  \n",
    "\n",
    "Filtering bad data. We'll need to check... \n",
    "1. gameevents have the same game list as games <span style=\"color:green\">- DONE</span>\n",
    "2. the game has finished <span style=\"color:green\"> - DONE</span><span style=\"color:orange\"> -  more naunced: dropouts will change the feature, \"no. of players\". Therefore \"no. of players\" will be assigned to the end no. of players (after dropout) </span>\n",
    "3. the map used (only use a sub-set of maps, or we could one-hot a list of accepted maps in X (features) <span style=\"color:green\">- DONE</span>\n",
    "4. additional rules? Look up <span style=\"color:orange\">- left for now as v bitty: will add a bit of noise but not a lot</span>\n",
    "5. Number of players. Possible for 2, 3, 4 or 5 player games, we could one-hot the differences as there might be similarities in different players that can contribute to predictions between different player games. Although >5 players should be removed. <span style=\"color:green\"> - DONE</span>\n",
    "6. No extra factions (shapeshifters etc)   <span style=\"color:orange\"> - more naunced: these will be filtered out at the stage of making the data, using \"valid factions\". However we don't want to filter out any games where a valid faction is playing against an extra faction (we just won't include that faction's vp data). So we include it in the \"player dropped\" group.</span>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = dict()\n",
    "data['gameevents'] = gameevents\n",
    "data['games'] = games\n",
    "data['gamescoringtiles'] = gamescoringtiles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "def filteringByBadgames(data, badgames):\n",
    "    \"\"\" Data is a dict containing gameevents, games, gamescoringtiles\n",
    "        badgames is a pd.dataframe that contains ['game'] to filter by\n",
    "    \"\"\"\n",
    "    gameeventsfil = data['gameevents']\n",
    "    gamesfil = data['games']\n",
    "    gamescoringtilesfil = data['gamescoringtiles']\n",
    "    \n",
    "    badgameslist = badgames['game']\n",
    "    gameeventsfilbefore = len(gameeventsfil)\n",
    "    gamesbefore = len(gamesfil)\n",
    "    gameSTbefore = len(gamescoringtilesfil)\n",
    "\n",
    "    gameeventsfil = gameeventsfil[~gameeventsfil['game'].isin(badgameslist)]\n",
    "    gamesfil = gamesfil[~gamesfil['game'].isin(badgameslist)]\n",
    "    gamescoringtilesfil = gamescoringtilesfil[~gamescoringtilesfil['game'].isin(badgameslist)]\n",
    "\n",
    "    print(f'game events before: {gameeventsfilbefore}, game events after: {len(gameeventsfil)}, game events removed: {gameeventsfilbefore-len(gameeventsfil)}')\n",
    "    print(f'games before: {gamesbefore}, games after: {len(gamesfil)}, games removed: {gamesbefore-len(gamesfil)}')\n",
    "    print(f'gameST before: {gameSTbefore}, gameST after: {len(gamescoringtilesfil)}, games removed: {gameSTbefore-len(gamescoringtilesfil)}')\n",
    "    \n",
    "    data['gameevents'] = gameeventsfil\n",
    "    data['games'] = gamesfil\n",
    "    data['gamescoringtiles'] = gamescoringtilesfil\n",
    "    \n",
    "    return data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### `Gameevents` games are same as `games` games"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# games from game_events are in games:\n",
    "gamesymmetricdif = set(gameslist) ^ set(games['game'])\n",
    "len(gamesymmetricdif)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So we have all the same games. Cool."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing the 1, 6, 7 player games\n",
    "This was a problem noted here: https://www.kaggle.com/lemonkoala/some-faulty-data"
   ]
  },
  {
   "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>base_map</th>\n",
       "      <th>game</th>\n",
       "      <th>last_update</th>\n",
       "      <th>player_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1267</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>HenriMustGoDown</td>\n",
       "      <td>2015-07-10 06:03:39</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3110</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>7pmatch3</td>\n",
       "      <td>2015-11-23 17:34:26</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3896</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>GTMDHXD</td>\n",
       "      <td>2015-11-02 05:54:08</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4760</th>\n",
       "      <td>95a66999127893f5925a5f591d54f8bcb9a670e6</td>\n",
       "      <td>seven11</td>\n",
       "      <td>2015-11-26 10:38:09</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5728</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>6playerGregvlwonthelastone</td>\n",
       "      <td>2016-01-11 14:43:19</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112011</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>7pmatch</td>\n",
       "      <td>2015-08-22 14:35:38</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112930</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>MEGA2KING</td>\n",
       "      <td>2015-08-26 12:21:04</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113411</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>ScruffyLookingNerfHerder</td>\n",
       "      <td>2015-08-18 02:13:18</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113813</th>\n",
       "      <td>be8f6ebf549404d015547152d5f2a1906ae8dd90</td>\n",
       "      <td>youmi023</td>\n",
       "      <td>2015-08-27 08:07:33</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113846</th>\n",
       "      <td>126fe960806d587c78546b30f1a90853b1ada468</td>\n",
       "      <td>YutoriGoGo009</td>\n",
       "      <td>2015-08-29 00:32:15</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>83 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        base_map                        game  \\\n",
       "1267    be8f6ebf549404d015547152d5f2a1906ae8dd90             HenriMustGoDown   \n",
       "3110    be8f6ebf549404d015547152d5f2a1906ae8dd90                    7pmatch3   \n",
       "3896    be8f6ebf549404d015547152d5f2a1906ae8dd90                     GTMDHXD   \n",
       "4760    95a66999127893f5925a5f591d54f8bcb9a670e6                     seven11   \n",
       "5728    be8f6ebf549404d015547152d5f2a1906ae8dd90  6playerGregvlwonthelastone   \n",
       "...                                          ...                         ...   \n",
       "112011  126fe960806d587c78546b30f1a90853b1ada468                     7pmatch   \n",
       "112930  be8f6ebf549404d015547152d5f2a1906ae8dd90                   MEGA2KING   \n",
       "113411  126fe960806d587c78546b30f1a90853b1ada468    ScruffyLookingNerfHerder   \n",
       "113813  be8f6ebf549404d015547152d5f2a1906ae8dd90                    youmi023   \n",
       "113846  126fe960806d587c78546b30f1a90853b1ada468               YutoriGoGo009   \n",
       "\n",
       "                last_update  player_count  \n",
       "1267    2015-07-10 06:03:39             6  \n",
       "3110    2015-11-23 17:34:26             7  \n",
       "3896    2015-11-02 05:54:08             7  \n",
       "4760    2015-11-26 10:38:09             7  \n",
       "5728    2016-01-11 14:43:19             6  \n",
       "...                     ...           ...  \n",
       "112011  2015-08-22 14:35:38             7  \n",
       "112930  2015-08-26 12:21:04             6  \n",
       "113411  2015-08-18 02:13:18             6  \n",
       "113813  2015-08-27 08:07:33             7  \n",
       "113846  2015-08-29 00:32:15             7  \n",
       "\n",
       "[83 rows x 4 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "badgames = games[games[\"player_count\"].isin([1, 6, 7])]\n",
    "badgames"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "game events before: 73419074, game events after: 73332500, game events removed: 86574\n",
      "games before: 115612, games after: 115529, games removed: 83\n",
      "gameST before: 693660, gameST after: 693162, games removed: 498\n"
     ]
    }
   ],
   "source": [
    "data = filteringByBadgames(data, badgames)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking the game has finished\n",
    "To check this, we can look at game events. If every player has passed their bonus tile on round 6, they've all made it to the end. If not, they've dropped out."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['action:ACT1', 'action:ACT2', 'action:ACT4', 'action:FAV6',\n",
       "       'advance:ship', 'bridge', 'build:D', 'burn', 'decline:count',\n",
       "       'decline:pw', 'dig', 'favor:FAV12', 'favor:any', 'leech:count',\n",
       "       'leech:pw', 'order:1', 'order:2', 'order:3', 'order:4', 'town:TW1',\n",
       "       'town:TW2', 'town:TW5', 'town:any', 'upgrade:TE', 'upgrade:TP',\n",
       "       'vp'], dtype=object)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "singlegame = gameevents[gameevents['game'] == 'NewbiesWelcome12']\n",
    "r6 = singlegame[singlegame['round'] == 6]\n",
    "uniqueevents = np.sort(pd.unique(r6['event']))\n",
    "uniqueevents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There's no pass logged in round 6?! What?! What about round 5?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['action:ACT1', 'action:ACT3', 'action:ACT6', 'action:ACTS',\n",
       "       'action:BON1', 'action:FAV6', 'bridge', 'build:D', 'burn', 'dig',\n",
       "       'favor:FAV1', 'favor:FAV10', 'favor:FAV5', 'favor:FAV6',\n",
       "       'favor:any', 'leech:count', 'leech:pw', 'order:1', 'order:2',\n",
       "       'order:3', 'order:4', 'pass:BON2', 'pass:BON4', 'pass:BON7',\n",
       "       'town:TW3', 'town:any', 'upgrade:SA', 'upgrade:TE', 'upgrade:TP',\n",
       "       'vp'], dtype=object)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "singlegame = gameevents[gameevents['game'] == 'NewbiesWelcome12']\n",
    "r5 = singlegame[singlegame['round'] == 5]\n",
    "uniqueevents = np.sort(pd.unique(r5['event']))\n",
    "uniqueevents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4388095    False\n",
       "4388096    False\n",
       "4388107    False\n",
       "4388109    False\n",
       "4388119    False\n",
       "           ...  \n",
       "4388605    False\n",
       "4388607    False\n",
       "4388612    False\n",
       "4388618    False\n",
       "4388619    False\n",
       "Name: event, Length: 92, dtype: bool"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r6['event'].str.startswith('pass')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Okay. So there is in round 5 but not in round 6. Not a fool-proof way of checking no-one has dropped out but you'd hope if they made it to round 6, they wouldn't drop out then. Very annoying, we'll just base it off round 5 passes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "faction: swarmlings ended their turn?: True\n",
      "faction: halflings ended their turn?: True\n",
      "faction: engineers ended their turn?: False\n",
      "faction: witches ended their turn?: True\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>nodrops</th>\n",
       "      <th>startplayers</th>\n",
       "      <th>playersdropped</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NewbiesWelcome12</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               game  nodrops  startplayers  playersdropped\n",
       "0  NewbiesWelcome12    False             4               1"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sense-check - every player passed final round except engineers in this game\n",
    "singlegame = gameevents[gameevents['game'] == 'NewbiesWelcome12']\n",
    "\n",
    "extendedfactions = validfactions + ['dragonlords', 'riverwalkers', 'yetis', 'icemaidens', 'shapeshifters', 'acolytes']\n",
    "\n",
    "def check_game_ended(singlegame, verbose=False):\n",
    "    r5 = singlegame[singlegame['round'] == 5]\n",
    "    rawfactions = pd.unique(singlegame['faction'])\n",
    "    verifiedfactions = [rawfaction for rawfaction in rawfactions if rawfaction in extendedfactions]\n",
    "    allbool = []\n",
    "\n",
    "    for faction in verifiedfactions:\n",
    "        factionbool = len(r5[(r5['faction'] == faction)  & (r5['event'].str.startswith('pass'))]) == 1\n",
    "        allbool.append(factionbool)\n",
    "        \n",
    "        if verbose:\n",
    "            print(f'faction: {faction} ended their turn?: {factionbool}')\n",
    "    \n",
    "    isgood = all(allbool)\n",
    "    startplayers = len(verifiedfactions)\n",
    "    boolsum = sum(allbool)\n",
    "    playersdropped = startplayers - boolsum\n",
    "    \n",
    "    return isgood, startplayers, playersdropped\n",
    "\n",
    "isgood, startplayers, playersdropped = check_game_ended(singlegame, verbose=True)\n",
    "newdf = pd.DataFrame([['NewbiesWelcome12', isgood, startplayers, playersdropped]], columns=['game', 'nodrops', 'startplayers', 'playersdropped'])\n",
    "newdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll also need to check there even *is* a round 5 (or a round 6). If there isn't... clearly the game hasn't been finished. We might as well check nothing odd is going on and there isn't a round 7/8/9 either."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Progressed to 10000th game\n",
      "Progressed to 20000th game\n",
      "Progressed to 30000th game\n",
      "Progressed to 40000th game\n",
      "Progressed to 50000th game\n",
      "Progressed to 60000th game\n",
      "Progressed to 70000th game\n",
      "Progressed to 80000th game\n",
      "Progressed to 90000th game\n",
      "Progressed to 100000th game\n",
      "Progressed to 110000th game\n",
      "no of unique games in table is: 115528\n"
     ]
    }
   ],
   "source": [
    "gameevents = data['gameevents']\n",
    "gameslist = list(pd.unique(gameevents['game']))\n",
    "\n",
    "gamelengthlen = len(gameslist)\n",
    "gamesroundup = math.ceil(gamelengthlen / 100.0) * 100\n",
    "jj = 0\n",
    "playerdropdf = pd.DataFrame(columns=['game', 'nodrops', 'startplayers', 'playersdropped'])\n",
    "\n",
    "for ii in range(100, gamesroundup+1, 100):     \n",
    "    ii = min(ii, gamelengthlen)    \n",
    "    if (ii % 10000) == 0:  # update every 10000 games\n",
    "        print(f'Progressed to {ii}th game')\n",
    "    \n",
    "    next100games = gameslist[jj:ii]\n",
    "    jj = ii  # so that we don't get any repetitions at the very end, where our set will be smaller\n",
    "    \n",
    "    gameevents100 = gameevents[gameevents['game'].isin(next100games)]\n",
    "\n",
    "    for game in next100games:\n",
    "        singlegame = gameevents100[gameevents100['game'] == game]\n",
    "        \n",
    "        if not len(singlegame) == 0:\n",
    "            isgood, startplayers, playersdropped = check_game_ended(singlegame)\n",
    "            newdf = pd.DataFrame([[game, isgood, startplayers, playersdropped]], columns=['game', 'nodrops', 'startplayers', 'playersdropped'])\n",
    "            playerdropdf = playerdropdf.append(newdf, ignore_index=True)\n",
    "\n",
    "print(f\"no of unique games in table is: {len(list(pd.unique(playerdropdf['game'])))}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    100147\n",
       "1     11975\n",
       "2      2720\n",
       "3       530\n",
       "4       133\n",
       "5        23\n",
       "Name: playersdropped, dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playerdropdf['playersdropped'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "playerdropdf['endplayers'] = playerdropdf['startplayers'] - playerdropdf['playersdropped']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4    57599\n",
       "3    25708\n",
       "2    23374\n",
       "5     6057\n",
       "1     1921\n",
       "0      869\n",
       "Name: endplayers, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playerdropdf['endplayers'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "playerdropdf.to_csv('D://PycharmProjects/TerraBot/data/faction-picker-bot/unfinishedgames.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for coming back to edit\n",
    "playerdropdf = pd.read_csv('D://PycharmProjects/TerraBot/data/faction-picker-bot/unfinishedgames.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Remove all the games which end up with 0 or 1 player."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>Unnamed: 0</th>\n",
       "      <th>game</th>\n",
       "      <th>nodrops</th>\n",
       "      <th>startplayers</th>\n",
       "      <th>playersdropped</th>\n",
       "      <th>endplayers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0000000001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>050615</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0512</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>051501</td>\n",
       "      <td>True</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>060303</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115523</th>\n",
       "      <td>115523</td>\n",
       "      <td>Youreacrookcaptianhook</td>\n",
       "      <td>False</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115524</th>\n",
       "      <td>115524</td>\n",
       "      <td>YourMomIsAChaosMagician</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115525</th>\n",
       "      <td>115525</td>\n",
       "      <td>ytuwertqwtr</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115526</th>\n",
       "      <td>115526</td>\n",
       "      <td>yuertyqert</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115527</th>\n",
       "      <td>115527</td>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>115528 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Unnamed: 0                     game  nodrops  startplayers  \\\n",
       "0                0               0000000001     True             2   \n",
       "1                1                   050615     True             4   \n",
       "2                2                     0512     True             4   \n",
       "3                3                   051501     True             5   \n",
       "4                4                   060303     True             4   \n",
       "...            ...                      ...      ...           ...   \n",
       "115523      115523   Youreacrookcaptianhook    False             3   \n",
       "115524      115524  YourMomIsAChaosMagician     True             3   \n",
       "115525      115525              ytuwertqwtr     True             2   \n",
       "115526      115526               yuertyqert     True             2   \n",
       "115527      115527        ZeicheMasZeuchnis     True             3   \n",
       "\n",
       "        playersdropped  endplayers  \n",
       "0                    0           2  \n",
       "1                    0           4  \n",
       "2                    0           4  \n",
       "3                    0           5  \n",
       "4                    0           4  \n",
       "...                ...         ...  \n",
       "115523               1           2  \n",
       "115524               0           3  \n",
       "115525               0           2  \n",
       "115526               0           2  \n",
       "115527               0           3  \n",
       "\n",
       "[115528 rows x 6 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playerdropdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "game events before: 73332500, game events after: 72827570, game events removed: 504930\n",
      "games before: 115529, games after: 112739, games removed: 2790\n",
      "gameST before: 693162, gameST after: 676434, games removed: 16728\n"
     ]
    }
   ],
   "source": [
    "badgames = playerdropdf[playerdropdf['endplayers'].isin([0, 1])]\n",
    "data = filteringByBadgames(data, badgames)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2790\n"
     ]
    }
   ],
   "source": [
    "print(len(playerdropdf[playerdropdf['endplayers'].isin([0, 1])]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing unwanted maps\n",
    "We want these maps:  \n",
    "126fe960806d587c78546b30f1a90853b1ada468 - map1  \n",
    "95a66999127893f5925a5f591d54f8bcb9a670e6 - map2  \n",
    "be8f6ebf549404d015547152d5f2a1906ae8dd90 - map3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "game events before: 72827570, game events after: 72067832, game events removed: 759738\n",
      "games before: 112739, games after: 111547, games removed: 1192\n",
      "gameST before: 676434, gameST after: 669282, games removed: 7152\n"
     ]
    }
   ],
   "source": [
    "acceptablemaps = ['126fe960806d587c78546b30f1a90853b1ada468',   \n",
    "                  '95a66999127893f5925a5f591d54f8bcb9a670e6', \n",
    "                  'be8f6ebf549404d015547152d5f2a1906ae8dd90']\n",
    "\n",
    "badgames = games[~games[\"base_map\"].isin(acceptablemaps)]\n",
    "data = filteringByBadgames(data, badgames)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating the full dataset <a name=\"fullDataset\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Progressed to 10000th game\n",
      "Progressed to 20000th game\n",
      "Progressed to 30000th game\n",
      "Progressed to 40000th game\n",
      "Progressed to 50000th game\n",
      "Progressed to 60000th game\n",
      "Progressed to 70000th game\n",
      "Progressed to 80000th game\n",
      "Progressed to 90000th game\n",
      "Progressed to 100000th game\n",
      "Progressed to 110000th game\n",
      "no of unique games in table is: 111546\n"
     ]
    }
   ],
   "source": [
    "vpdf, _, _ = makenewdf()\n",
    "featdf, _ = emptyfeaturesdf()\n",
    "\n",
    "gameevents = data['gameevents']\n",
    "games = data['games']\n",
    "gamescoringtiles = data['gamescoringtiles']\n",
    "\n",
    "gameslist = list(pd.unique(gameevents['game']))\n",
    "gamelengthlen = len(gameslist)\n",
    "gamesroundup = math.ceil(gamelengthlen / 100.0) * 100\n",
    "jj = 0\n",
    "\n",
    "for ii in range(100, gamesroundup+1, 100):     \n",
    "    ii = min(ii, gamelengthlen)    \n",
    "    if (ii % 10000) == 0:  # update every 10000 games\n",
    "        print(f'Progressed to {ii}th game')\n",
    "    \n",
    "    next100games = gameslist[jj:ii]\n",
    "    jj = ii  # so that we don't get any repetitions at the very end, where our set will be smaller\n",
    "    \n",
    "    gameevents100 = gameevents[gameevents['game'].isin(next100games)]\n",
    "    gamemeta100 = games[games['game'].isin(next100games)]\n",
    "    gameST100 = gamescoringtiles[gamescoringtiles['game'].isin(next100games)]\n",
    "    endplayers100 = playerdropdf[playerdropdf['game'].isin(next100games)]  # use this for player count\n",
    "\n",
    "    for game in next100games:\n",
    "        singlegame = gameevents100[gameevents100['game'] == game]\n",
    "        singlegamemeta = gamemeta100[gamemeta100['game'] == game]\n",
    "        singlegameST = gameST100[gameST100['game'] == game]\n",
    "        singleendplayers = endplayers100[endplayers100['game'] == game]\n",
    "        \n",
    "        if not len(singlegame) == 0:\n",
    "            vpforgame = get_vp_from_game(singlegame)\n",
    "            featsforgame = get_features_from_game(singlegame, singlegamemeta, singlegameST, singleendplayers=singleendplayers)\n",
    "            \n",
    "            vpdf = vpdf.append(vpforgame, ignore_index=True)\n",
    "            featdf = featdf.append(featsforgame, ignore_index=True)\n",
    "\n",
    "print(f\"no of unique games in table is: {len(list(pd.unique(vpdf['game'])))}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>game</th>\n",
       "      <th>witches</th>\n",
       "      <th>auren</th>\n",
       "      <th>swarmlings</th>\n",
       "      <th>mermaids</th>\n",
       "      <th>cultists</th>\n",
       "      <th>halflings</th>\n",
       "      <th>dwarves</th>\n",
       "      <th>engineers</th>\n",
       "      <th>chaosmagicians</th>\n",
       "      <th>giants</th>\n",
       "      <th>fakirs</th>\n",
       "      <th>nomads</th>\n",
       "      <th>darklings</th>\n",
       "      <th>alchemists</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>118.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>62.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>050615</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>122.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>124.0</td>\n",
       "      <td>101.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>051501</td>\n",
       "      <td>NaN</td>\n",
       "      <td>84.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>77.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>72.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>54.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>060303</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>108.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>121.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>78.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>128.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\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",
       "      <th>111541</th>\n",
       "      <td>Youreacrookcaptianhook</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>126.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>105.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>37.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111542</th>\n",
       "      <td>YourMomIsAChaosMagician</td>\n",
       "      <td>112.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>114.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>120.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111543</th>\n",
       "      <td>ytuwertqwtr</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>105.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>102.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111544</th>\n",
       "      <td>yuertyqert</td>\n",
       "      <td>128.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>141.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111545</th>\n",
       "      <td>ZeicheMasZeuchnis</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>132.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>88.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111546 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           game  witches  auren  swarmlings  mermaids  \\\n",
       "0                    0000000001      NaN    NaN         NaN     118.0   \n",
       "1                        050615      NaN    NaN         NaN       NaN   \n",
       "2                          0512      NaN    NaN         NaN       NaN   \n",
       "3                        051501      NaN   84.0         NaN       NaN   \n",
       "4                        060303      NaN    NaN         NaN     108.0   \n",
       "...                         ...      ...    ...         ...       ...   \n",
       "111541   Youreacrookcaptianhook      NaN    NaN       126.0       NaN   \n",
       "111542  YourMomIsAChaosMagician    112.0    NaN       114.0       NaN   \n",
       "111543              ytuwertqwtr      NaN    NaN       105.0       NaN   \n",
       "111544               yuertyqert    128.0    NaN         NaN       NaN   \n",
       "111545        ZeicheMasZeuchnis      NaN    NaN         NaN       NaN   \n",
       "\n",
       "        cultists  halflings  dwarves  engineers  chaosmagicians  giants  \\\n",
       "0            NaN       62.0      NaN        NaN             NaN     NaN   \n",
       "1            NaN        NaN      NaN        NaN             NaN     NaN   \n",
       "2            NaN        NaN      NaN      124.0           101.0     NaN   \n",
       "3            NaN       77.0      NaN        NaN            72.0     NaN   \n",
       "4            NaN        NaN      NaN      121.0             NaN     NaN   \n",
       "...          ...        ...      ...        ...             ...     ...   \n",
       "111541       NaN      105.0      NaN        NaN             NaN     NaN   \n",
       "111542       NaN        NaN      NaN        NaN             NaN     NaN   \n",
       "111543       NaN        NaN    102.0        NaN             NaN     NaN   \n",
       "111544       NaN        NaN      NaN        NaN           141.0     NaN   \n",
       "111545       NaN        NaN    132.0        NaN             NaN     NaN   \n",
       "\n",
       "        fakirs  nomads  darklings  alchemists  \n",
       "0          NaN     NaN        NaN         NaN  \n",
       "1          NaN     NaN      122.0         NaN  \n",
       "2          NaN     NaN        NaN         NaN  \n",
       "3          NaN    54.0       78.0         NaN  \n",
       "4         78.0     NaN      128.0         NaN  \n",
       "...        ...     ...        ...         ...  \n",
       "111541     NaN    37.0        NaN         NaN  \n",
       "111542     NaN   120.0        NaN         NaN  \n",
       "111543     NaN     NaN        NaN         NaN  \n",
       "111544     NaN     NaN        NaN         NaN  \n",
       "111545     NaN    88.0      120.0         NaN  \n",
       "\n",
       "[111546 rows x 15 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vpdf"
   ]
  },
  {
   "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>game</th>\n",
       "      <th>r1_SCORE1</th>\n",
       "      <th>r1_SCORE2</th>\n",
       "      <th>r1_SCORE3</th>\n",
       "      <th>r1_SCORE4</th>\n",
       "      <th>r1_SCORE5</th>\n",
       "      <th>r1_SCORE6</th>\n",
       "      <th>r1_SCORE7</th>\n",
       "      <th>r1_SCORE8</th>\n",
       "      <th>r1_SCORE9</th>\n",
       "      <th>...</th>\n",
       "      <th>BON8</th>\n",
       "      <th>BON9</th>\n",
       "      <th>BON10</th>\n",
       "      <th>2players</th>\n",
       "      <th>3players</th>\n",
       "      <th>4players</th>\n",
       "      <th>5players</th>\n",
       "      <th>map1</th>\n",
       "      <th>map2</th>\n",
       "      <th>map3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0000000001</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>050615</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0512</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>051501</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>060303</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111541</th>\n",
       "      <td>Youreacrookcaptianhook</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111542</th>\n",
       "      <td>YourMomIsAChaosMagician</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111543</th>\n",
       "      <td>ytuwertqwtr</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111544</th>\n",
       "      <td>yuertyqert</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111545</th>\n",
       "      <td>ZeicheMasZeuchnis</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111546 rows × 72 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           game r1_SCORE1 r1_SCORE2 r1_SCORE3 r1_SCORE4  \\\n",
       "0                    0000000001         0         0         0         0   \n",
       "1                        050615         0         0         0         0   \n",
       "2                          0512         0         0         0         0   \n",
       "3                        051501         0         0         0         0   \n",
       "4                        060303         0         1         0         0   \n",
       "...                         ...       ...       ...       ...       ...   \n",
       "111541   Youreacrookcaptianhook         0         0         0         0   \n",
       "111542  YourMomIsAChaosMagician         0         0         1         0   \n",
       "111543              ytuwertqwtr         1         0         0         0   \n",
       "111544               yuertyqert         0         0         0         1   \n",
       "111545        ZeicheMasZeuchnis         0         0         1         0   \n",
       "\n",
       "       r1_SCORE5 r1_SCORE6 r1_SCORE7 r1_SCORE8 r1_SCORE9  ... BON8 BON9 BON10  \\\n",
       "0              0         0         0         1         0  ...    0    1     1   \n",
       "1              1         0         0         0         0  ...    1    1     0   \n",
       "2              1         0         0         0         0  ...    1    1     0   \n",
       "3              0         0         0         1         0  ...    1    0     0   \n",
       "4              0         0         0         0         0  ...    1    1     1   \n",
       "...          ...       ...       ...       ...       ...  ...  ...  ...   ...   \n",
       "111541         0         1         0         0         0  ...    0    1     0   \n",
       "111542         0         0         0         0         0  ...    1    1     0   \n",
       "111543         0         0         0         0         0  ...    1    1     1   \n",
       "111544         0         0         0         0         0  ...    1    0     1   \n",
       "111545         0         0         0         0         0  ...    1    1     0   \n",
       "\n",
       "       2players 3players 4players 5players map1 map2 map3  \n",
       "0             1        0        0        0    1    0    0  \n",
       "1             0        0        1        0    0    0    1  \n",
       "2             0        0        1        0    1    0    0  \n",
       "3             0        0        0        1    1    0    0  \n",
       "4             0        0        1        0    0    1    0  \n",
       "...         ...      ...      ...      ...  ...  ...  ...  \n",
       "111541        1        0        0        0    1    0    0  \n",
       "111542        0        1        0        0    1    0    0  \n",
       "111543        1        0        0        0    1    0    0  \n",
       "111544        1        0        0        0    1    0    0  \n",
       "111545        0        1        0        0    1    0    0  \n",
       "\n",
       "[111546 rows x 72 columns]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "featdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "vpdf.to_csv('D://PycharmProjects/TerraBot/data/faction-picker-bot/vpdata.csv')\n",
    "featdf.to_csv('D://PycharmProjects/TerraBot/data/faction-picker-bot/featdata.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check we still have the same games:\n",
    "gamesymmetricdif = set(vpdf['game']) ^ set(featdf['game'])\n",
    "len(gamesymmetricdif)"
   ]
  }
 ],
 "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": 4
}