{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We will follow the methodology mentioned in the report which is as follows:\n",
"\n",
"<ol>\n",
" <li>Find out the frame of highest speed within each game for each player</li>\n",
" <li>Find the corresponding maximum values of accelLoad and accelImpulse if there are multiple frames with highest speed</li>\n",
" <li>Give Speed, AccelLoad and AccelImpulse equal weights and scale the score to 100.</li>\n",
"</ol>"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(4570160, 14)"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"# Caution: The file is about 500 MB in size, so it'll take around 2-3 min to load\n",
"performance_file = 'https://www.dropbox.com/s/n7pvlxy60qwyy91/gps.csv?dl=1'\n",
"performance = pd.read_csv(performance_file)\n",
"performance.shape"
]
},
{
"cell_type": "code",
"execution_count": 2,
"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>GameID</th>\n",
" <th>Half</th>\n",
" <th>PlayerID</th>\n",
" <th>FrameID</th>\n",
" <th>Time</th>\n",
" <th>GameClock</th>\n",
" <th>Speed</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>AccelX</th>\n",
" <th>AccelY</th>\n",
" <th>AccelZ</th>\n",
" <th>Longitude</th>\n",
" <th>Latitude</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>4843</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>4844</td>\n",
" <td>00:30:05</td>\n",
" <td>00:08:04</td>\n",
" <td>6.752783</td>\n",
" <td>3.138891</td>\n",
" <td>0.168341</td>\n",
" <td>-0.29250</td>\n",
" <td>1.83250</td>\n",
" <td>1.50000</td>\n",
" <td>55.466833</td>\n",
" <td>24.994974</td>\n",
" </tr>\n",
" <tr>\n",
" <td>7763</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>2507</td>\n",
" <td>00:26:11</td>\n",
" <td>00:04:10</td>\n",
" <td>7.277784</td>\n",
" <td>0.333334</td>\n",
" <td>0.177188</td>\n",
" <td>0.18500</td>\n",
" <td>3.49875</td>\n",
" <td>0.67500</td>\n",
" <td>55.466410</td>\n",
" <td>24.995237</td>\n",
" </tr>\n",
" <tr>\n",
" <td>11005</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>4</td>\n",
" <td>492</td>\n",
" <td>00:22:50</td>\n",
" <td>00:00:49</td>\n",
" <td>7.638895</td>\n",
" <td>0.361111</td>\n",
" <td>0.112861</td>\n",
" <td>0.57125</td>\n",
" <td>0.75750</td>\n",
" <td>0.44875</td>\n",
" <td>55.466054</td>\n",
" <td>24.995193</td>\n",
" </tr>\n",
" <tr>\n",
" <td>20654</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>6</td>\n",
" <td>4884</td>\n",
" <td>00:30:09</td>\n",
" <td>00:08:08</td>\n",
" <td>5.191671</td>\n",
" <td>3.055558</td>\n",
" <td>0.121655</td>\n",
" <td>0.56000</td>\n",
" <td>2.19125</td>\n",
" <td>1.38750</td>\n",
" <td>55.466794</td>\n",
" <td>24.994953</td>\n",
" </tr>\n",
" <tr>\n",
" <td>23511</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>7</td>\n",
" <td>2484</td>\n",
" <td>00:26:09</td>\n",
" <td>00:04:08</td>\n",
" <td>6.577783</td>\n",
" <td>0.000000</td>\n",
" <td>0.035796</td>\n",
" <td>0.45500</td>\n",
" <td>1.00875</td>\n",
" <td>0.61250</td>\n",
" <td>55.466563</td>\n",
" <td>24.995140</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" GameID Half PlayerID FrameID Time GameClock Speed \\\n",
"4843 1 1 2 4844 00:30:05 00:08:04 6.752783 \n",
"7763 1 1 3 2507 00:26:11 00:04:10 7.277784 \n",
"11005 1 1 4 492 00:22:50 00:00:49 7.638895 \n",
"20654 1 1 6 4884 00:30:09 00:08:08 5.191671 \n",
"23511 1 1 7 2484 00:26:09 00:04:08 6.577783 \n",
"\n",
" AccelImpulse AccelLoad AccelX AccelY AccelZ Longitude \\\n",
"4843 3.138891 0.168341 -0.29250 1.83250 1.50000 55.466833 \n",
"7763 0.333334 0.177188 0.18500 3.49875 0.67500 55.466410 \n",
"11005 0.361111 0.112861 0.57125 0.75750 0.44875 55.466054 \n",
"20654 3.055558 0.121655 0.56000 2.19125 1.38750 55.466794 \n",
"23511 0.000000 0.035796 0.45500 1.00875 0.61250 55.466563 \n",
"\n",
" Latitude \n",
"4843 24.994974 \n",
"7763 24.995237 \n",
"11005 24.995193 \n",
"20654 24.994953 \n",
"23511 24.995140 "
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_performance = performance.loc[performance['Speed'] != 0]\n",
"idx = df_performance.groupby(['GameID','Half','PlayerID'])['Speed'].transform(max) == df_performance['Speed']\n",
"df_max_speed = df_performance[idx]\n",
"df_max_speed.head()"
]
},
{
"cell_type": "code",
"execution_count": 3,
"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>GameID</th>\n",
" <th>Half</th>\n",
" <th>PlayerID</th>\n",
" <th>Speed</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>6.752783</td>\n",
" <td>3.138891</td>\n",
" <td>0.168341</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>7.277784</td>\n",
" <td>0.333334</td>\n",
" <td>0.177188</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>4</td>\n",
" <td>7.638895</td>\n",
" <td>0.361111</td>\n",
" <td>0.112861</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>6</td>\n",
" <td>5.191671</td>\n",
" <td>3.055558</td>\n",
" <td>0.121655</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>7</td>\n",
" <td>6.577783</td>\n",
" <td>0.000000</td>\n",
" <td>0.035796</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>989</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>13</td>\n",
" <td>7.086117</td>\n",
" <td>4.194448</td>\n",
" <td>0.150259</td>\n",
" </tr>\n",
" <tr>\n",
" <td>990</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>14</td>\n",
" <td>7.433339</td>\n",
" <td>3.888892</td>\n",
" <td>0.078690</td>\n",
" </tr>\n",
" <tr>\n",
" <td>991</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>15</td>\n",
" <td>4.736115</td>\n",
" <td>2.222224</td>\n",
" <td>0.015441</td>\n",
" </tr>\n",
" <tr>\n",
" <td>992</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>16</td>\n",
" <td>6.158338</td>\n",
" <td>0.000000</td>\n",
" <td>0.116188</td>\n",
" </tr>\n",
" <tr>\n",
" <td>993</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>16</td>\n",
" <td>6.158338</td>\n",
" <td>0.000000</td>\n",
" <td>0.111365</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>994 rows × 6 columns</p>\n",
"</div>"
],
"text/plain": [
" GameID Half PlayerID Speed AccelImpulse AccelLoad\n",
"0 1 1 2 6.752783 3.138891 0.168341\n",
"1 1 1 3 7.277784 0.333334 0.177188\n",
"2 1 1 4 7.638895 0.361111 0.112861\n",
"3 1 1 6 5.191671 3.055558 0.121655\n",
"4 1 1 7 6.577783 0.000000 0.035796\n",
".. ... ... ... ... ... ...\n",
"989 38 2 13 7.086117 4.194448 0.150259\n",
"990 38 2 14 7.433339 3.888892 0.078690\n",
"991 38 2 15 4.736115 2.222224 0.015441\n",
"992 38 2 16 6.158338 0.000000 0.116188\n",
"993 38 2 16 6.158338 0.000000 0.111365\n",
"\n",
"[994 rows x 6 columns]"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_max_speed = df_max_speed[['GameID','Half','PlayerID','Speed','AccelImpulse','AccelLoad']]\n",
"index = df_max_speed.groupby(['GameID','Half','PlayerID'])['AccelImpulse'].transform(max) == df_max_speed['AccelImpulse']\n",
"df_final_performance = df_max_speed[index]\n",
"df_final_performance.reset_index(inplace=True,drop=True)\n",
"df_final_performance"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As we don't have the Date of the games in our data, we'll import that information from the games dataset"
]
},
{
"cell_type": "code",
"execution_count": 4,
"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>GameID</th>\n",
" <th>Date</th>\n",
" <th>Tournament</th>\n",
" <th>TournamentGame</th>\n",
" <th>Team</th>\n",
" <th>Opponent</th>\n",
" <th>Outcome</th>\n",
" <th>TeamPoints</th>\n",
" <th>TeamPointsAllowed</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>Dubai</td>\n",
" <td>1</td>\n",
" <td>Canada</td>\n",
" <td>Spain</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>Dubai</td>\n",
" <td>2</td>\n",
" <td>Canada</td>\n",
" <td>Ireland</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>Dubai</td>\n",
" <td>3</td>\n",
" <td>Canada</td>\n",
" <td>Fiji</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" <td>14</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>4</td>\n",
" <td>2017-12-01</td>\n",
" <td>Dubai</td>\n",
" <td>4</td>\n",
" <td>Canada</td>\n",
" <td>France</td>\n",
" <td>W</td>\n",
" <td>24</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>5</td>\n",
" <td>2017-12-01</td>\n",
" <td>Dubai</td>\n",
" <td>5</td>\n",
" <td>Canada</td>\n",
" <td>Australia</td>\n",
" <td>L</td>\n",
" <td>7</td>\n",
" <td>25</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" GameID Date Tournament TournamentGame Team Opponent Outcome \\\n",
"0 1 2017-11-30 Dubai 1 Canada Spain W \n",
"1 2 2017-11-30 Dubai 2 Canada Ireland W \n",
"2 3 2017-11-30 Dubai 3 Canada Fiji W \n",
"3 4 2017-12-01 Dubai 4 Canada France W \n",
"4 5 2017-12-01 Dubai 5 Canada Australia L \n",
"\n",
" TeamPoints TeamPointsAllowed \n",
"0 19 0 \n",
"1 31 0 \n",
"2 31 14 \n",
"3 24 19 \n",
"4 7 25 "
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"game_file = 'https://www.dropbox.com/s/tk13ad7sca5nkwy/games.csv?dl=1'\n",
"games = pd.read_csv(game_file)\n",
"games.head()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"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>GameID</th>\n",
" <th>Half</th>\n",
" <th>PlayerID</th>\n",
" <th>Speed</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Date</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>2</td>\n",
" <td>6.752783</td>\n",
" <td>3.138891</td>\n",
" <td>0.168341</td>\n",
" <td>2017-11-30</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>3</td>\n",
" <td>7.277784</td>\n",
" <td>0.333334</td>\n",
" <td>0.177188</td>\n",
" <td>2017-11-30</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>4</td>\n",
" <td>7.638895</td>\n",
" <td>0.361111</td>\n",
" <td>0.112861</td>\n",
" <td>2017-11-30</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>6</td>\n",
" <td>5.191671</td>\n",
" <td>3.055558</td>\n",
" <td>0.121655</td>\n",
" <td>2017-11-30</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>1</td>\n",
" <td>1</td>\n",
" <td>7</td>\n",
" <td>6.577783</td>\n",
" <td>0.000000</td>\n",
" <td>0.035796</td>\n",
" <td>2017-11-30</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>989</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>13</td>\n",
" <td>7.086117</td>\n",
" <td>4.194448</td>\n",
" <td>0.150259</td>\n",
" <td>2018-07-21</td>\n",
" </tr>\n",
" <tr>\n",
" <td>990</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>14</td>\n",
" <td>7.433339</td>\n",
" <td>3.888892</td>\n",
" <td>0.078690</td>\n",
" <td>2018-07-21</td>\n",
" </tr>\n",
" <tr>\n",
" <td>991</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>15</td>\n",
" <td>4.736115</td>\n",
" <td>2.222224</td>\n",
" <td>0.015441</td>\n",
" <td>2018-07-21</td>\n",
" </tr>\n",
" <tr>\n",
" <td>992</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>16</td>\n",
" <td>6.158338</td>\n",
" <td>0.000000</td>\n",
" <td>0.116188</td>\n",
" <td>2018-07-21</td>\n",
" </tr>\n",
" <tr>\n",
" <td>993</td>\n",
" <td>38</td>\n",
" <td>2</td>\n",
" <td>16</td>\n",
" <td>6.158338</td>\n",
" <td>0.000000</td>\n",
" <td>0.111365</td>\n",
" <td>2018-07-21</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>994 rows × 7 columns</p>\n",
"</div>"
],
"text/plain": [
" GameID Half PlayerID Speed AccelImpulse AccelLoad Date\n",
"0 1 1 2 6.752783 3.138891 0.168341 2017-11-30\n",
"1 1 1 3 7.277784 0.333334 0.177188 2017-11-30\n",
"2 1 1 4 7.638895 0.361111 0.112861 2017-11-30\n",
"3 1 1 6 5.191671 3.055558 0.121655 2017-11-30\n",
"4 1 1 7 6.577783 0.000000 0.035796 2017-11-30\n",
".. ... ... ... ... ... ... ...\n",
"989 38 2 13 7.086117 4.194448 0.150259 2018-07-21\n",
"990 38 2 14 7.433339 3.888892 0.078690 2018-07-21\n",
"991 38 2 15 4.736115 2.222224 0.015441 2018-07-21\n",
"992 38 2 16 6.158338 0.000000 0.116188 2018-07-21\n",
"993 38 2 16 6.158338 0.000000 0.111365 2018-07-21\n",
"\n",
"[994 rows x 7 columns]"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_games = games[['GameID','Date']]\n",
"dfPerformance = pd.merge(df_final_performance, df_games, left_on='GameID',right_on='GameID', how='left')\n",
"dfPerformance"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now, as we have the wellness and training load information only for each day, we'll have to come up with a measure of performance of each player per day rather than per game. Therefore, we groupby the Date of games and take the average of all games on the same date."
]
},
{
"cell_type": "code",
"execution_count": 6,
"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>Date</th>\n",
" <th>PlayerID</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Speed</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>2.761907</td>\n",
" <td>0.091576</td>\n",
" <td>6.857942</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>2.392363</td>\n",
" <td>0.112402</td>\n",
" <td>6.851742</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>4</td>\n",
" <td>2.202383</td>\n",
" <td>0.107841</td>\n",
" <td>6.404370</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>6</td>\n",
" <td>3.060188</td>\n",
" <td>0.098006</td>\n",
" <td>6.234727</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-11-30</td>\n",
" <td>7</td>\n",
" <td>3.527781</td>\n",
" <td>0.069148</td>\n",
" <td>5.503824</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>207</td>\n",
" <td>2018-07-21</td>\n",
" <td>11</td>\n",
" <td>1.883335</td>\n",
" <td>0.104770</td>\n",
" <td>7.061117</td>\n",
" </tr>\n",
" <tr>\n",
" <td>208</td>\n",
" <td>2018-07-21</td>\n",
" <td>13</td>\n",
" <td>2.597224</td>\n",
" <td>0.111778</td>\n",
" <td>6.369450</td>\n",
" </tr>\n",
" <tr>\n",
" <td>209</td>\n",
" <td>2018-07-21</td>\n",
" <td>14</td>\n",
" <td>3.800003</td>\n",
" <td>0.064752</td>\n",
" <td>4.713893</td>\n",
" </tr>\n",
" <tr>\n",
" <td>210</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>2.597224</td>\n",
" <td>0.088764</td>\n",
" <td>5.272226</td>\n",
" </tr>\n",
" <tr>\n",
" <td>211</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.958334</td>\n",
" <td>0.082387</td>\n",
" <td>5.023152</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>212 rows × 5 columns</p>\n",
"</div>"
],
"text/plain": [
" Date PlayerID AccelImpulse AccelLoad Speed\n",
"0 2017-11-30 2 2.761907 0.091576 6.857942\n",
"1 2017-11-30 3 2.392363 0.112402 6.851742\n",
"2 2017-11-30 4 2.202383 0.107841 6.404370\n",
"3 2017-11-30 6 3.060188 0.098006 6.234727\n",
"4 2017-11-30 7 3.527781 0.069148 5.503824\n",
".. ... ... ... ... ...\n",
"207 2018-07-21 11 1.883335 0.104770 7.061117\n",
"208 2018-07-21 13 2.597224 0.111778 6.369450\n",
"209 2018-07-21 14 3.800003 0.064752 4.713893\n",
"210 2018-07-21 15 2.597224 0.088764 5.272226\n",
"211 2018-07-21 16 0.958334 0.082387 5.023152\n",
"\n",
"[212 rows x 5 columns]"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dfPer = dfPerformance[dfPerformance.columns.difference([\"GameID\",\"Half\"])]\n",
"dfPer = dfPer.groupby(['Date','PlayerID']).mean()\n",
"dfPer.reset_index(inplace=True)\n",
"dfPer"
]
},
{
"cell_type": "code",
"execution_count": 7,
"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>Date</th>\n",
" <th>PlayerID</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Speed</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>4</td>\n",
" <td>0.415584</td>\n",
" <td>0.318376</td>\n",
" <td>0.685600</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>6</td>\n",
" <td>0.582358</td>\n",
" <td>0.285724</td>\n",
" <td>0.659945</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-11-30</td>\n",
" <td>7</td>\n",
" <td>0.673267</td>\n",
" <td>0.189919</td>\n",
" <td>0.549412</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>207</td>\n",
" <td>2018-07-21</td>\n",
" <td>11</td>\n",
" <td>0.353555</td>\n",
" <td>0.308181</td>\n",
" <td>0.784919</td>\n",
" </tr>\n",
" <tr>\n",
" <td>208</td>\n",
" <td>2018-07-21</td>\n",
" <td>13</td>\n",
" <td>0.492349</td>\n",
" <td>0.331446</td>\n",
" <td>0.680319</td>\n",
" </tr>\n",
" <tr>\n",
" <td>209</td>\n",
" <td>2018-07-21</td>\n",
" <td>14</td>\n",
" <td>0.726193</td>\n",
" <td>0.175328</td>\n",
" <td>0.429952</td>\n",
" </tr>\n",
" <tr>\n",
" <td>210</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" </tr>\n",
" <tr>\n",
" <td>211</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>212 rows × 5 columns</p>\n",
"</div>"
],
"text/plain": [
" Date PlayerID AccelImpulse AccelLoad Speed\n",
"0 2017-11-30 2 0.524367 0.264378 0.754193\n",
"1 2017-11-30 3 0.452520 0.333518 0.753256\n",
"2 2017-11-30 4 0.415584 0.318376 0.685600\n",
"3 2017-11-30 6 0.582358 0.285724 0.659945\n",
"4 2017-11-30 7 0.673267 0.189919 0.549412\n",
".. ... ... ... ... ...\n",
"207 2018-07-21 11 0.353555 0.308181 0.784919\n",
"208 2018-07-21 13 0.492349 0.331446 0.680319\n",
"209 2018-07-21 14 0.726193 0.175328 0.429952\n",
"210 2018-07-21 15 0.492349 0.255042 0.514388\n",
"211 2018-07-21 16 0.173717 0.233874 0.476721\n",
"\n",
"[212 rows x 5 columns]"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from sklearn import preprocessing\n",
"# Scale the three values in the range 0-1 first\n",
"min_max_scaler = preprocessing.MinMaxScaler()\n",
"dfPer[[\"Speed\",\"AccelImpulse\",\"AccelLoad\"]] = min_max_scaler.fit_transform(dfPer[[\"Speed\",\"AccelImpulse\",\"AccelLoad\"]])\n",
"dfPer"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>Date</th>\n",
" <th>PlayerID</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Speed</th>\n",
" <th>PerformanceScore</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" <td>51.309794</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>4</td>\n",
" <td>0.415584</td>\n",
" <td>0.318376</td>\n",
" <td>0.685600</td>\n",
" <td>47.318679</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>6</td>\n",
" <td>0.582358</td>\n",
" <td>0.285724</td>\n",
" <td>0.659945</td>\n",
" <td>50.934243</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-11-30</td>\n",
" <td>7</td>\n",
" <td>0.673267</td>\n",
" <td>0.189919</td>\n",
" <td>0.549412</td>\n",
" <td>47.086623</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>207</td>\n",
" <td>2018-07-21</td>\n",
" <td>11</td>\n",
" <td>0.353555</td>\n",
" <td>0.308181</td>\n",
" <td>0.784919</td>\n",
" <td>48.221864</td>\n",
" </tr>\n",
" <tr>\n",
" <td>208</td>\n",
" <td>2018-07-21</td>\n",
" <td>13</td>\n",
" <td>0.492349</td>\n",
" <td>0.331446</td>\n",
" <td>0.680319</td>\n",
" <td>50.137139</td>\n",
" </tr>\n",
" <tr>\n",
" <td>209</td>\n",
" <td>2018-07-21</td>\n",
" <td>14</td>\n",
" <td>0.726193</td>\n",
" <td>0.175328</td>\n",
" <td>0.429952</td>\n",
" <td>44.382406</td>\n",
" </tr>\n",
" <tr>\n",
" <td>210</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" <td>42.059287</td>\n",
" </tr>\n",
" <tr>\n",
" <td>211</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" <td>29.477054</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>212 rows × 6 columns</p>\n",
"</div>"
],
"text/plain": [
" Date PlayerID AccelImpulse AccelLoad Speed PerformanceScore\n",
"0 2017-11-30 2 0.524367 0.264378 0.754193 51.431257\n",
"1 2017-11-30 3 0.452520 0.333518 0.753256 51.309794\n",
"2 2017-11-30 4 0.415584 0.318376 0.685600 47.318679\n",
"3 2017-11-30 6 0.582358 0.285724 0.659945 50.934243\n",
"4 2017-11-30 7 0.673267 0.189919 0.549412 47.086623\n",
".. ... ... ... ... ... ...\n",
"207 2018-07-21 11 0.353555 0.308181 0.784919 48.221864\n",
"208 2018-07-21 13 0.492349 0.331446 0.680319 50.137139\n",
"209 2018-07-21 14 0.726193 0.175328 0.429952 44.382406\n",
"210 2018-07-21 15 0.492349 0.255042 0.514388 42.059287\n",
"211 2018-07-21 16 0.173717 0.233874 0.476721 29.477054\n",
"\n",
"[212 rows x 6 columns]"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Calculate the PerformanceScore on the scale 0-100\n",
"dfPer[\"PerformanceScore\"] = (dfPer[\"AccelImpulse\"] + dfPer[\"AccelLoad\"] + dfPer[\"Speed\"])*100/3\n",
"dfPer"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We also did some simple preprocessing to find out the points difference and the outcome of each game, to see if these values have significance with respect to the PerformanceScore. "
]
},
{
"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>Date</th>\n",
" <th>GameID</th>\n",
" <th>Outcome</th>\n",
" <th>PointsDiff</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>1</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>W</td>\n",
" <td>17</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-12-01</td>\n",
" <td>4</td>\n",
" <td>W</td>\n",
" <td>5</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-12-01</td>\n",
" <td>5</td>\n",
" <td>L</td>\n",
" <td>-18</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Date GameID Outcome PointsDiff\n",
"0 2017-11-30 1 W 19\n",
"1 2017-11-30 2 W 31\n",
"2 2017-11-30 3 W 17\n",
"3 2017-12-01 4 W 5\n",
"4 2017-12-01 5 L -18"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"games = pd.read_csv('https://raw.githubusercontent.com/weilixiang/sta2453_project1/master/Clean%20Data/games.csv')\n",
"games.head()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"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>Date</th>\n",
" <th>PlayerID</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Speed</th>\n",
" <th>PerformanceScore</th>\n",
" <th>GameID</th>\n",
" <th>Outcome</th>\n",
" <th>PointsDiff</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>1</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>2</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>3</td>\n",
" <td>W</td>\n",
" <td>17</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" <td>51.309794</td>\n",
" <td>1</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" <td>51.309794</td>\n",
" <td>2</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>477</td>\n",
" <td>2018-07-21</td>\n",
" <td>14</td>\n",
" <td>0.726193</td>\n",
" <td>0.175328</td>\n",
" <td>0.429952</td>\n",
" <td>44.382406</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>478</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" <td>42.059287</td>\n",
" <td>37</td>\n",
" <td>L</td>\n",
" <td>-12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>479</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" <td>42.059287</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>480</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" <td>29.477054</td>\n",
" <td>37</td>\n",
" <td>L</td>\n",
" <td>-12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>481</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" <td>29.477054</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>482 rows × 9 columns</p>\n",
"</div>"
],
"text/plain": [
" Date PlayerID AccelImpulse AccelLoad Speed \\\n",
"0 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"1 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"2 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"3 2017-11-30 3 0.452520 0.333518 0.753256 \n",
"4 2017-11-30 3 0.452520 0.333518 0.753256 \n",
".. ... ... ... ... ... \n",
"477 2018-07-21 14 0.726193 0.175328 0.429952 \n",
"478 2018-07-21 15 0.492349 0.255042 0.514388 \n",
"479 2018-07-21 15 0.492349 0.255042 0.514388 \n",
"480 2018-07-21 16 0.173717 0.233874 0.476721 \n",
"481 2018-07-21 16 0.173717 0.233874 0.476721 \n",
"\n",
" PerformanceScore GameID Outcome PointsDiff \n",
"0 51.431257 1 W 19 \n",
"1 51.431257 2 W 31 \n",
"2 51.431257 3 W 17 \n",
"3 51.309794 1 W 19 \n",
"4 51.309794 2 W 31 \n",
".. ... ... ... ... \n",
"477 44.382406 38 W 12 \n",
"478 42.059287 37 L -12 \n",
"479 42.059287 38 W 12 \n",
"480 29.477054 37 L -12 \n",
"481 29.477054 38 W 12 \n",
"\n",
"[482 rows x 9 columns]"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dfPer = pd.merge(dfPer, games, left_on='Date',right_on='Date', how='left')\n",
"dfPer"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To see if our PerformanceScore shows variability with respect to players and games, let us visualize the data for the first 5 players."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1080x504 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"from matplotlib import pyplot as plt\n",
"fig, ax = plt.subplots(figsize=(15,7))\n",
"plt.title('Player Performance Comparison')\n",
"firstfive = dfPer[\"PlayerID\"].isin([1,2,3,4,5])\n",
"dfPerfive = dfPer[firstfive]\n",
"dfPerfive.set_index('GameID', inplace=True)\n",
"dfPerfive.groupby(['PlayerID'])['PerformanceScore'].plot(legend='True')\n",
"plt.ylabel(\"Performance Score\")\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"80.75302804382518"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dfPer[\"PerformanceScore\"].max()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"17.016373365206878"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dfPer[\"PerformanceScore\"].min()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As we can see, our PerformanceScore has a good range of values. You can also notice that now games on the same day will have the same PerformanceScore per player which makes it easier for us to analyze the data with respect to training load and wellness. Finally, we'll remove players 18-21 from our dataset as they hadn't played any games according to the information given on the project website. "
]
},
{
"cell_type": "code",
"execution_count": 15,
"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>Date</th>\n",
" <th>PlayerID</th>\n",
" <th>AccelImpulse</th>\n",
" <th>AccelLoad</th>\n",
" <th>Speed</th>\n",
" <th>PerformanceScore</th>\n",
" <th>GameID</th>\n",
" <th>Outcome</th>\n",
" <th>PointsDiff</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td>0</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>1</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>1</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>2</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <td>2</td>\n",
" <td>2017-11-30</td>\n",
" <td>2</td>\n",
" <td>0.524367</td>\n",
" <td>0.264378</td>\n",
" <td>0.754193</td>\n",
" <td>51.431257</td>\n",
" <td>3</td>\n",
" <td>W</td>\n",
" <td>17</td>\n",
" </tr>\n",
" <tr>\n",
" <td>3</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" <td>51.309794</td>\n",
" <td>1</td>\n",
" <td>W</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <td>4</td>\n",
" <td>2017-11-30</td>\n",
" <td>3</td>\n",
" <td>0.452520</td>\n",
" <td>0.333518</td>\n",
" <td>0.753256</td>\n",
" <td>51.309794</td>\n",
" <td>2</td>\n",
" <td>W</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <td>462</td>\n",
" <td>2018-07-21</td>\n",
" <td>14</td>\n",
" <td>0.726193</td>\n",
" <td>0.175328</td>\n",
" <td>0.429952</td>\n",
" <td>44.382406</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>463</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" <td>42.059287</td>\n",
" <td>37</td>\n",
" <td>L</td>\n",
" <td>-12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>464</td>\n",
" <td>2018-07-21</td>\n",
" <td>15</td>\n",
" <td>0.492349</td>\n",
" <td>0.255042</td>\n",
" <td>0.514388</td>\n",
" <td>42.059287</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>465</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" <td>29.477054</td>\n",
" <td>37</td>\n",
" <td>L</td>\n",
" <td>-12</td>\n",
" </tr>\n",
" <tr>\n",
" <td>466</td>\n",
" <td>2018-07-21</td>\n",
" <td>16</td>\n",
" <td>0.173717</td>\n",
" <td>0.233874</td>\n",
" <td>0.476721</td>\n",
" <td>29.477054</td>\n",
" <td>38</td>\n",
" <td>W</td>\n",
" <td>12</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>467 rows × 9 columns</p>\n",
"</div>"
],
"text/plain": [
" Date PlayerID AccelImpulse AccelLoad Speed \\\n",
"0 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"1 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"2 2017-11-30 2 0.524367 0.264378 0.754193 \n",
"3 2017-11-30 3 0.452520 0.333518 0.753256 \n",
"4 2017-11-30 3 0.452520 0.333518 0.753256 \n",
".. ... ... ... ... ... \n",
"462 2018-07-21 14 0.726193 0.175328 0.429952 \n",
"463 2018-07-21 15 0.492349 0.255042 0.514388 \n",
"464 2018-07-21 15 0.492349 0.255042 0.514388 \n",
"465 2018-07-21 16 0.173717 0.233874 0.476721 \n",
"466 2018-07-21 16 0.173717 0.233874 0.476721 \n",
"\n",
" PerformanceScore GameID Outcome PointsDiff \n",
"0 51.431257 1 W 19 \n",
"1 51.431257 2 W 31 \n",
"2 51.431257 3 W 17 \n",
"3 51.309794 1 W 19 \n",
"4 51.309794 2 W 31 \n",
".. ... ... ... ... \n",
"462 44.382406 38 W 12 \n",
"463 42.059287 37 L -12 \n",
"464 42.059287 38 W 12 \n",
"465 29.477054 37 L -12 \n",
"466 29.477054 38 W 12 \n",
"\n",
"[467 rows x 9 columns]"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"dfPer = dfPer.drop(dfPer[(dfPer['PlayerID']>17)].index)\n",
"dfPer.reset_index(inplace=True,drop=True)\n",
"dfPer"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"dfPer.to_csv(\"Clean Data/PerformanceMeasure.csv\", index = None, header=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"<b>Bonus</b>:You can run the cell below to generate a cool GIF of Player wise PerformanceScore over Games! :)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from celluloid import Camera\n",
"\n",
"fig = plt.figure()\n",
"camera = Camera(fig)\n",
"plt.xlabel('Games')\n",
"plt.ylabel('Performance Score')\n",
"plt.title('Player Performance over time (Game) GIF')\n",
"dfPer.set_index('GameID', inplace=True)\n",
"for i in range(17):\n",
" t = plt.plot(dfPer[dfPer[\"PlayerID\"] == i+1].PerformanceScore)\n",
" plt.legend(t, [f'Player {i+1}'])\n",
" camera.snap()\n",
"animation = camera.animate()\n",
"animation.save('Performance.gif', writer = 'imagemagick')\n",
"dfPer.reset_index(inplace=True)\n",
"dfPer.head()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.3"
},
"toc": {
"base_numbering": 1,
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": true,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": true,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": false
}
},
"nbformat": 4,
"nbformat_minor": 2
}