Among_Us_Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d0bb6035",
   "metadata": {},
   "source": [
    "# Data Analysis and Prediction on Among Us Dataset\n",
    "#### Fall 2021 DS-1001 Capstone Project\n",
    "#### Xu Han xh852, Jason Wang jw7383, Chloe Zheng cz1300"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac004b57",
   "metadata": {},
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5d343cda",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2fe17c50",
   "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 Completed Date</th>\n",
       "      <th>Team</th>\n",
       "      <th>Outcome</th>\n",
       "      <th>Task Completed</th>\n",
       "      <th>All Tasks Completed</th>\n",
       "      <th>Murdered</th>\n",
       "      <th>Imposter Kills</th>\n",
       "      <th>Game Length</th>\n",
       "      <th>Ejected</th>\n",
       "      <th>Sabotages Fixed</th>\n",
       "      <th>Time to complete all tasks</th>\n",
       "      <th>Rank Change</th>\n",
       "      <th>Region/Game Code</th>\n",
       "      <th>User</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/13/2020 at 1:26:56 am EST</td>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Win</td>\n",
       "      <td>3</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>-</td>\n",
       "      <td>07m 04s</td>\n",
       "      <td>No</td>\n",
       "      <td>2.0</td>\n",
       "      <td>-</td>\n",
       "      <td>++</td>\n",
       "      <td>NA / WYMSBF</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/13/2020 at 1:17:42 am EST</td>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Loss</td>\n",
       "      <td>7</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>-</td>\n",
       "      <td>16m 21s</td>\n",
       "      <td>No</td>\n",
       "      <td>1.0</td>\n",
       "      <td>09m 48s</td>\n",
       "      <td>--</td>\n",
       "      <td>NA / WYMSBF</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/13/2020 at 12:57:47 am EST</td>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Win</td>\n",
       "      <td>3</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>-</td>\n",
       "      <td>11m 33s</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-</td>\n",
       "      <td>++</td>\n",
       "      <td>NA / WYMSBF</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/13/2020 at 12:41:55 am EST</td>\n",
       "      <td>Imposter</td>\n",
       "      <td>Win</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>2</td>\n",
       "      <td>08m 05s</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-</td>\n",
       "      <td>+++</td>\n",
       "      <td>Europe / QIRTNF</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/13/2020 at 12:30:37 am EST</td>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Loss</td>\n",
       "      <td>4</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>-</td>\n",
       "      <td>05m 10s</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-</td>\n",
       "      <td>---</td>\n",
       "      <td>Europe / QIRTNF</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Game Completed Date      Team Outcome Task Completed  \\\n",
       "0   12/13/2020 at 1:26:56 am EST  Crewmate     Win              3   \n",
       "1   12/13/2020 at 1:17:42 am EST  Crewmate    Loss              7   \n",
       "2  12/13/2020 at 12:57:47 am EST  Crewmate     Win              3   \n",
       "3  12/13/2020 at 12:41:55 am EST  Imposter     Win              -   \n",
       "4  12/13/2020 at 12:30:37 am EST  Crewmate    Loss              4   \n",
       "\n",
       "  All Tasks Completed Murdered Imposter Kills Game Length Ejected  \\\n",
       "0                  No      Yes              -     07m 04s      No   \n",
       "1                 Yes       No              -     16m 21s      No   \n",
       "2                  No       No              -     11m 33s      No   \n",
       "3                   -        -              2     08m 05s      No   \n",
       "4                  No       No              -     05m 10s      No   \n",
       "\n",
       "   Sabotages Fixed Time to complete all tasks Rank Change Region/Game Code  \\\n",
       "0              2.0                          -          ++      NA / WYMSBF   \n",
       "1              1.0                    09m 48s          --      NA / WYMSBF   \n",
       "2              0.0                          -          ++      NA / WYMSBF   \n",
       "3              NaN                          -         +++  Europe / QIRTNF   \n",
       "4              0.0                          -         ---  Europe / QIRTNF   \n",
       "\n",
       "   User  \n",
       "0     0  \n",
       "1     0  \n",
       "2     0  \n",
       "3     0  \n",
       "4     0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import Data and concatenate into one Dataframe\n",
    "# Add User Id\n",
    "files = []\n",
    "for i in range(29):\n",
    "    tempdf = pd.read_csv(f'data/User{i+1}.csv', delimiter = ',')\n",
    "    tempdf['User'] = i\n",
    "    files.append(tempdf)\n",
    "df = pd.concat(files, ignore_index = True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e7e5213a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Game Completed Date            object\n",
       "Team                           object\n",
       "Outcome                        object\n",
       "Task Completed                 object\n",
       "All Tasks Completed            object\n",
       "Murdered                       object\n",
       "Imposter Kills                 object\n",
       "Game Length                    object\n",
       "Ejected                        object\n",
       "Sabotages Fixed               float64\n",
       "Time to complete all tasks     object\n",
       "Rank Change                    object\n",
       "Region/Game Code               object\n",
       "User                            int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List column features\n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6780b271",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reformat Data - time and region\n",
    "df[['Task Completed', 'Imposter Kills']] = df[['Task Completed', 'Imposter Kills']].apply(pd.to_numeric, errors = 'coerce')\n",
    "df['Region'] = df['Region/Game Code'].str.extract(r'^(.*?)/')\n",
    "df['Region'].value_counts()\n",
    "df['Game Length'] = df['Game Length'].apply(pd.to_timedelta)\n",
    "df['Game Length'] = df['Game Length'] / np.timedelta64(1, 's')\n",
    "df['Time to complete all tasks'] = df['Time to complete all tasks'].apply(pd.to_timedelta, errors = 'coerce')\n",
    "df['Time to complete all tasks'] = df['Time to complete all tasks'] / np.timedelta64(1, 's')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "39fb0ed5",
   "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>Team</th>\n",
       "      <th>Outcome</th>\n",
       "      <th>Task Completed</th>\n",
       "      <th>All Tasks Completed</th>\n",
       "      <th>Murdered</th>\n",
       "      <th>Game Length</th>\n",
       "      <th>Ejected</th>\n",
       "      <th>Sabotages Fixed</th>\n",
       "      <th>Region</th>\n",
       "      <th>User</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Win</td>\n",
       "      <td>3.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>424.0</td>\n",
       "      <td>No</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NA</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Loss</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>981.0</td>\n",
       "      <td>No</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NA</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Win</td>\n",
       "      <td>3.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>693.0</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NA</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Loss</td>\n",
       "      <td>4.0</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>310.0</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Europe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>Loss</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>982.0</td>\n",
       "      <td>No</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Europe</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Team Outcome  Task Completed All Tasks Completed Murdered  Game Length  \\\n",
       "0  Crewmate     Win             3.0                  No      Yes        424.0   \n",
       "1  Crewmate    Loss             7.0                 Yes       No        981.0   \n",
       "2  Crewmate     Win             3.0                  No       No        693.0   \n",
       "4  Crewmate    Loss             4.0                  No       No        310.0   \n",
       "5  Crewmate    Loss             7.0                 Yes      Yes        982.0   \n",
       "\n",
       "  Ejected  Sabotages Fixed   Region  User  \n",
       "0      No              2.0      NA      0  \n",
       "1      No              1.0      NA      0  \n",
       "2      No              0.0      NA      0  \n",
       "4      No              0.0  Europe      0  \n",
       "5      No              0.0  Europe      0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create Dataframe for ONLY Crewmate data\n",
    "crewmate = df[['Team',\n",
    "               'Outcome',\n",
    "               'Task Completed',\n",
    "               'All Tasks Completed',\n",
    "               'Murdered',\n",
    "               'Game Length',\n",
    "               'Ejected',\n",
    "              'Sabotages Fixed',\n",
    "              'Region',\n",
    "              'User']][df['Team'] == 'Crewmate']\n",
    "crewmate.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b4cb1fe9",
   "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>Team</th>\n",
       "      <th>Outcome</th>\n",
       "      <th>Task Completed</th>\n",
       "      <th>All Tasks Completed</th>\n",
       "      <th>Murdered</th>\n",
       "      <th>Game Length</th>\n",
       "      <th>Ejected</th>\n",
       "      <th>Sabotages Fixed</th>\n",
       "      <th>Region</th>\n",
       "      <th>User</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>424.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>981.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>693.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>310.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>982.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Team  Outcome  Task Completed  All Tasks Completed  Murdered  \\\n",
       "0  Crewmate        1             3.0                    0         1   \n",
       "1  Crewmate        0             7.0                    1         0   \n",
       "2  Crewmate        1             3.0                    0         0   \n",
       "4  Crewmate        0             4.0                    0         0   \n",
       "5  Crewmate        0             7.0                    1         1   \n",
       "\n",
       "   Game Length  Ejected  Sabotages Fixed  Region  User  \n",
       "0        424.0        0              2.0       0     0  \n",
       "1        981.0        0              1.0       0     0  \n",
       "2        693.0        0              0.0       0     0  \n",
       "4        310.0        0              0.0       1     0  \n",
       "5        982.0        0              0.0       1     0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reformat Data - change columns into binary\n",
    "crewmate['Outcome'] = crewmate['Outcome'].replace(['Loss', 'Win'],[0, 1])\n",
    "crewmate['All Tasks Completed'] = crewmate['All Tasks Completed'].replace(['No', 'Yes'],[0, 1])\n",
    "crewmate['Murdered'] = crewmate['Murdered'].replace(['No', 'Yes'],[0, 1])\n",
    "crewmate['Ejected'] = crewmate['Ejected'].replace(['No', 'Yes'],[0, 1])\n",
    "crewmate['Region'] = crewmate['Region'].replace(['NA ', 'Europe '],[0, 1])\n",
    "crewmate.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "591d5c6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Analyze Correlation Matrix for correlated features\n",
    "%matplotlib inline\n",
    "corrlation_matrix = crewmate.corr()\n",
    "fig, ax = plt.subplots(figsize = (20, 15))\n",
    "ax = sns.heatmap(corrlation_matrix,\n",
    "                annot = True,\n",
    "                linewidths = 0.5,\n",
    "                fmt = \".2f\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7abbd3f1",
   "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>Team</th>\n",
       "      <th>Outcome</th>\n",
       "      <th>Task Completed</th>\n",
       "      <th>Murdered</th>\n",
       "      <th>Game Length</th>\n",
       "      <th>Ejected</th>\n",
       "      <th>Sabotages Fixed</th>\n",
       "      <th>Region</th>\n",
       "      <th>User</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>424.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>981.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>693.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>310.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Crewmate</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>982.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Team  Outcome  Task Completed  Murdered  Game Length  Ejected  \\\n",
       "0  Crewmate        1             3.0         1        424.0        0   \n",
       "1  Crewmate        0             7.0         0        981.0        0   \n",
       "2  Crewmate        1             3.0         0        693.0        0   \n",
       "4  Crewmate        0             4.0         0        310.0        0   \n",
       "5  Crewmate        0             7.0         1        982.0        0   \n",
       "\n",
       "   Sabotages Fixed  Region  User  \n",
       "0              2.0       0     0  \n",
       "1              1.0       0     0  \n",
       "2              0.0       0     0  \n",
       "4              0.0       1     0  \n",
       "5              0.0       1     0  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Based on Correlation Matrix, 'All Tasks Completed' is highly correlated with 'Tasks Completed'\n",
    "#Drop 'All Tasks Completed' to resolve multicollinarity\n",
    "crewmate = crewmate.drop(['All Tasks Completed'], axis =1)\n",
    "crewmate.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4eb9881",
   "metadata": {},
   "source": [
    "## Hypothesis Testing\n",
    "\n",
    "#### Null Hypothesis: Crewmates who finish more tasks do not win more often."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6d2f4140",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "from scipy.stats import t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "07212e4b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p-value: 0.05237999155491819\n",
      "t-statistic: -1.6786449499637606\n"
     ]
    }
   ],
   "source": [
    "# t-test\n",
    "avg_task_completed = crewmate.groupby(by = 'User')['Task Completed'].mean()\n",
    "avg_wins = crewmate.groupby(by = 'User')['Outcome'].mean()\n",
    "\n",
    "thresh = avg_task_completed.mean() #Setting a threshold to separate data into two groups \n",
    "less_tasks = avg_wins[avg_task_completed < thresh]\n",
    "greater_tasks = avg_wins[avg_task_completed >= thresh]\n",
    "\n",
    "t1, p1 = stats.ttest_ind(less_tasks, greater_tasks, alternative='less') #Calculate t-statstic and p-value\n",
    "print(\"p-value:\", p1)\n",
    "print(\"t-statistic:\", t1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "cf090684",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Graph t-distribution with our results\n",
    "degree = 27\n",
    "fig = plt.figure()\n",
    "x = np.linspace(t.ppf(0.00001, degree),\n",
    "                t.ppf(0.99999, degree), 100)\n",
    "plt.plot(x, t.pdf(x, degree),'r-', alpha=0.6, label='t pdf')\n",
    "plt.axvline(t1, color = 'black', ls = '--')\n",
    "plt.xlabel('t')\n",
    "plt.ylabel('pdf')\n",
    "plt.title(\"pdf of Student's t-test\")\n",
    "plt.xlim(-5,5)\n",
    "plt.ylim(0,.4)\n",
    "\n",
    "def f(x):\n",
    "    return t.pdf(x, 27)\n",
    "\n",
    "section = np.linspace(-5,t1,100)\n",
    "plt.fill_between(section,f(section))\n",
    "plt.annotate('t=-1.678', (-3.2,.3))\n",
    "plt.annotate('p=0.052', (-3.6,.05))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdaba3c9",
   "metadata": {},
   "source": [
    "## Logistic Regression\n",
    "\n",
    "#### Predictors: Crewmate stats -(1) task completed (2) murdered (3) game length (4) ejected (5) sabotages fixed (6) region\n",
    "#### Outcome: Win or loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "616ca47d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "d7641026",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train/Test split functions\n",
    "# Goal: split data such that each unique user's data is ONLY in either the train or test.\n",
    "#     Since the amount of games played per user is NOT consistent, use the subset sum dynamic programming method to\n",
    "#     find a subset that satifies the target ratio\n",
    "#     REFERENCE: https://levelup.gitconnected.com/dynamic-programming-subset-sum-c386126621cd\n",
    "\n",
    "def user_train_valid_split(users_i,valid_ratio,possible=False):\n",
    "    n = sum(users_i)\n",
    "    target = int(np.floor(valid_ratio*n))\n",
    "    while not possible:\n",
    "        possible, validation_indices = user_split(users_i,target)\n",
    "        target -= 1\n",
    "    training_indices = [i for i in range(len(users_i)) if i not in validation_indices]\n",
    "    return training_indices, validation_indices\n",
    "\n",
    "def user_split(users_i,target):  \n",
    "    n = len(users_i)\n",
    "    solution = [[False for j in range(int(target+1))] for i in range(n+1)]\n",
    "    # base cases\n",
    "    for i in range(n):\n",
    "        solution[i][0] = True\n",
    "    # other cases\n",
    "    for i in range(1,n+1):\n",
    "        for j in range(1,target+1):\n",
    "            solution[i][j] = solution[i-1][j]\n",
    "            if(solution[i][j] == False and j >= users_i[i-1]):\n",
    "                solution[i][j] = solution[i][j] or solution[i-1][j-users_i[i-1]]\n",
    "    # check if the subset sum is possible\n",
    "    possible = solution[len(users_i)][target]\n",
    "    subset = []\n",
    "    if not possible: return possible, subset\n",
    "    # return the subset solution if one exists\n",
    "    y = len(users_i)\n",
    "    x = target\n",
    "    while x != 0:\n",
    "        if solution[y-1][x] == False:\n",
    "            subset.append(y-1)\n",
    "            x -= users_i[y-1]\n",
    "        else: \n",
    "            y -= 1\n",
    "    return possible, subset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "718372bd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "expected ratio: 0.25, actual: 0.2498580352072686\n"
     ]
    }
   ],
   "source": [
    "# Split data\n",
    "crewmates_i = list(crewmate[\"User\"].value_counts(sort = False))\n",
    "test_ratio = .25\n",
    "training_indices, test_indices = user_train_valid_split(crewmates_i,test_ratio)\n",
    "\n",
    "train_data = crewmate.loc[crewmate['User'].isin(training_indices)]\n",
    "test_data = crewmate.loc[crewmate['User'].isin(test_indices)]\n",
    "train_data.reset_index(drop=True, inplace=True)\n",
    "test_data.reset_index(drop=True, inplace=True)\n",
    "\n",
    "# Check\n",
    "train_sum = sum([crewmates_i[i] for i in training_indices])\n",
    "test_sum = sum([crewmates_i[i] for i in test_indices])\n",
    "print(\"expected ratio: {}, actual: {}\".format(test_ratio, test_sum/(train_sum+test_sum)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5377cca0",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training data...\n",
      "Testing model...\n",
      "AUC score: 0.6136017272034544\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Train data\n",
    "print(\"Training data...\")\n",
    "X_train = train_data.drop([\"Outcome\", \"Team\", \"User\"], axis=1)\n",
    "y_train = train_data[\"Outcome\"]\n",
    "clf = LogisticRegression(random_state=42).fit(X_train, y_train)\n",
    "coefs = clf.coef_[0]\n",
    "\n",
    "# Test data\n",
    "print(\"Testing model...\")\n",
    "X_test = test_data.drop([\"Outcome\", \"Team\", \"User\"], axis=1)\n",
    "y_true = test_data[\"Outcome\"]\n",
    "y_predict = clf.predict(X_test)\n",
    "\n",
    "auc = roc_auc_score(y_true, y_predict)\n",
    "print(\"AUC score:\", auc)\n",
    "\n",
    "# Analyze coefficients\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.bar(X_train.columns, coefs)\n",
    "plt.xlabel(\"Features\")\n",
    "plt.ylabel(\"Coeffcient\")\n",
    "plt.title(\"Predictor Coefficients\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c190286e",
   "metadata": {},
   "source": [
    "## Multiple Linear Regression\n",
    "\n",
    "#### Predictors: Grouped by user, crewmate stats - (1) avg winrate (2) avg number of task completed (3) number of times being crewmate (4) avg murdered rate (5) avg ejected rate (6) avg sabotages fixed (7) avg game length\n",
    "#### Outcome: Grouped by user, imposter win rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "dfefe7b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.linear_model import Lasso, Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "34b8c34d",
   "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>avg_winrate</th>\n",
       "      <th>avg_task_comp</th>\n",
       "      <th>num_time_crew</th>\n",
       "      <th>avg_murdered_rate</th>\n",
       "      <th>avg_ejected_rate</th>\n",
       "      <th>avg_sabotaged_fixed</th>\n",
       "      <th>avg_game_length</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>User</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.571429</td>\n",
       "      <td>6.142857</td>\n",
       "      <td>77</td>\n",
       "      <td>0.597403</td>\n",
       "      <td>0.012987</td>\n",
       "      <td>0.389610</td>\n",
       "      <td>741.363636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.480519</td>\n",
       "      <td>5.818182</td>\n",
       "      <td>77</td>\n",
       "      <td>0.467532</td>\n",
       "      <td>0.051948</td>\n",
       "      <td>0.675325</td>\n",
       "      <td>636.623377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.485294</td>\n",
       "      <td>5.352941</td>\n",
       "      <td>68</td>\n",
       "      <td>0.426471</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.661765</td>\n",
       "      <td>683.514706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.493827</td>\n",
       "      <td>4.135802</td>\n",
       "      <td>81</td>\n",
       "      <td>0.419753</td>\n",
       "      <td>0.061728</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>457.790123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.483871</td>\n",
       "      <td>4.806452</td>\n",
       "      <td>31</td>\n",
       "      <td>0.580645</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.451613</td>\n",
       "      <td>708.935484</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      avg_winrate  avg_task_comp  num_time_crew  avg_murdered_rate  \\\n",
       "User                                                                 \n",
       "0        0.571429       6.142857             77           0.597403   \n",
       "1        0.480519       5.818182             77           0.467532   \n",
       "2        0.485294       5.352941             68           0.426471   \n",
       "3        0.493827       4.135802             81           0.419753   \n",
       "4        0.483871       4.806452             31           0.580645   \n",
       "\n",
       "      avg_ejected_rate  avg_sabotaged_fixed  avg_game_length  \n",
       "User                                                          \n",
       "0             0.012987             0.389610       741.363636  \n",
       "1             0.051948             0.675325       636.623377  \n",
       "2             0.029412             0.661765       683.514706  \n",
       "3             0.061728             0.444444       457.790123  \n",
       "4             0.000000             0.451613       708.935484  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Prepare X data\n",
    "users = pd.DataFrame()\n",
    "\n",
    "users['avg_winrate'] = crewmate.groupby('User')['Outcome'].mean()\n",
    "users['avg_task_comp'] = crewmate.groupby('User')['Task Completed'].mean()\n",
    "users['num_time_crew'] = crewmate['User'].value_counts(sort=False)\n",
    "users['avg_murdered_rate'] = crewmate.groupby('User')['Murdered'].mean()\n",
    "users['avg_ejected_rate'] = crewmate.groupby('User')['Ejected'].mean()\n",
    "users['avg_sabotaged_fixed'] = crewmate.groupby('User')['Sabotages Fixed'].mean()\n",
    "users['avg_game_length'] = crewmate.groupby('User')['Game Length'].mean()\n",
    "\n",
    "users.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "7db7b132",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Prepare y data\n",
    "imp = df[df['Team']=='Imposter']\n",
    "imp['Outcome'] = imp['Outcome'].replace(['Loss', 'Win'], [0,1])\n",
    "target = imp.groupby('User')['Outcome'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9affb7c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data using built in function\n",
    "X_train, X_test, y_train, y_test = train_test_split(users, \n",
    "                                                    target, \n",
    "                                                    test_size=0.2,\n",
    "                                                   random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8a529fcd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alphas: [0.1, 1, 10, 100, 1000, 10000]\n",
      "train error: [0.014062169510723976, 0.017047862926834297, 0.018277474833601558, 0.019586431290974923, 0.019990264530667243, 0.02025152153290846]\n",
      "test error: [0.031329027183491655, 0.025455190357740923, 0.02407066862424419, 0.023656652997387576, 0.023640890671047743, 0.02383709743361236]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Ridge\n",
    "alphas = [0.1, 1, 10, 100, 1000,10000]\n",
    "train_error, test_error = [], []\n",
    "\n",
    "for alpha in alphas:\n",
    "    clf = Ridge(alpha=alpha)\n",
    "    clf.fit(X_train, y_train)\n",
    "    pers_train_predict = clf.predict(X_train)\n",
    "    pers_test_predict = clf.predict(X_test)\n",
    "    train_error.append(mean_squared_error(y_train, pers_train_predict))\n",
    "    test_error.append(mean_squared_error(y_test,pers_test_predict ))\n",
    "    \n",
    "print(\"alphas: {}\\ntrain error: {}\\ntest error: {}\".format(alphas,train_error,test_error))\n",
    "\n",
    "plt.plot(alphas, train_error, '-+', label='train')\n",
    "plt.plot(alphas, test_error,'-+', label='test')\n",
    "plt.xscale('log')\n",
    "plt.xlabel('alpha')\n",
    "plt.ylabel('mean squared error')\n",
    "plt.title('Ridge Regression: alpha vs. mse')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "bcb7b13e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "alphas: [0.001, 0.01, 0.1, 1]\n",
      "train error: [0.013351083045980117, 0.01816712966506992, 0.020055568714751525, 0.02111504369772632]\n",
      "test error: [0.034129538966012986, 0.024198806614650865, 0.023721080135123904, 0.02512092489759861]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Lasso\n",
    "lbds = [1e-3, 1e-2, 1e-1, 1]\n",
    "train_error_lasso, test_error_lasso = [], []\n",
    "\n",
    "for lbd in lbds:\n",
    "    lasso = Lasso(alpha=lbd)\n",
    "    lasso.fit(X_train, y_train)\n",
    "    \n",
    "    pers_train_predict = lasso.predict(X_train)\n",
    "    pers_test_predict = lasso.predict(X_test)\n",
    "    train_error_lasso.append(mean_squared_error(y_train, pers_train_predict))\n",
    "    test_error_lasso.append(mean_squared_error(y_test,pers_test_predict ))\n",
    "\n",
    "print(\"alphas: {}\\ntrain error: {}\\ntest error: {}\".format(lbds,train_error_lasso,test_error_lasso))\n",
    "\n",
    "plt.plot(lbds, train_error_lasso, '-+', label='train')\n",
    "plt.plot(lbds, test_error_lasso,'-+', label='test')\n",
    "plt.xscale('log')\n",
    "plt.xlabel('lambda')\n",
    "plt.ylabel('mean squared error')\n",
    "plt.title('Lasso Regression: lambda vs. mse')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "26ad8861",
   "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>outcome</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>User</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.563920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.812500</td>\n",
       "      <td>0.554246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.522264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.576287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.515077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.541667</td>\n",
       "      <td>0.525964</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       outcome  predicted\n",
       "User                     \n",
       "27    0.600000   0.563920\n",
       "16    0.812500   0.554246\n",
       "12    0.750000   0.522264\n",
       "22    0.714286   0.576287\n",
       "8     0.571429   0.515077\n",
       "9     0.541667   0.525964"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "#Update Lasso with best parameter 0.1\n",
    "best_lasso = Lasso(alpha = 0.1)\n",
    "#Fit the dataset into the new model\n",
    "best_lasso.fit(X_train, y_train)\n",
    "#Make predictions\n",
    "predict = best_lasso.predict(X_test)\n",
    "#Put real outcomes and predictions together\n",
    "lasso_result = pd.DataFrame()\n",
    "lasso_result['outcome'] = y_test\n",
    "lasso_result['predicted'] = predict\n",
    "lasso_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "06332008",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.019550787408172118"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculating mean squared error between y_test and the mean of imposter win rate \n",
    "avg_imp_winrate = np.zeros(6)\n",
    "avg_imp_winrate.fill(target.mean())\n",
    "mean_squared_error(avg_imp_winrate, y_test)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}