/
TimeseriesMerging.py
112 lines (74 loc) · 4.25 KB
/
TimeseriesMerging.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import os
import joblib
from utils import string2json
# from config import TIMESTEP
import argparse
import sys
plt.interactive(True)
pd.options.display.max_columns = 15
pic_prefix = 'pic/'
data_dict_resampled = joblib.load('data/data_dict_resampled')
gamedata_dict = joblib.load('data/gamedata_dict')
parser = argparse.ArgumentParser()
parser.add_argument('--TIMESTEP', default=10, type=float)
if __debug__:
print('SUPER WARNING!!! YOU ARE INTO DEBUG MODE', file=sys.stderr)
args = parser.parse_args(['--TIMESTEP=10'])
else:
args = parser.parse_args()
TIMESTEP = args.TIMESTEP
data_dict_resampled_merged = {}
def timestamp2step(times, df_start_time):
return np.round((np.array(times) - df_start_time) / TIMESTEP).astype(int)
# player_id = list(gamedata_dict.keys())[0] # DEBUG
for player_id in gamedata_dict:
df_resampled4player = data_dict_resampled[player_id]
df_resampled4player = df_resampled4player.reset_index()
gamedata_dict4player = gamedata_dict[player_id]
time_game_start = pd.to_datetime(gamedata_dict4player['time_game_start'], unit='s')
time_game_end = pd.to_datetime(gamedata_dict4player['time_game_end'], unit='s')
mask_gametime = (time_game_start < df_resampled4player['time']) & (df_resampled4player['time'] < time_game_end)
df_resampled4player = df_resampled4player.loc[mask_gametime] # Only data during the game is now used
df_resampled4player['time'] = df_resampled4player['time'].apply(lambda x: x.timestamp())
df_start_time = df_resampled4player['time'].min()
times_kills = timestamp2step(gamedata_dict4player['times_kills'], df_start_time) # TODO: check. I found a bug
times_deaths = timestamp2step(gamedata_dict4player['times_is_killed'], df_start_time) # TODO: check. I found a bug
if 'shootout_times_start_end' in gamedata_dict4player:
times_shootouts = timestamp2step(gamedata_dict4player['shootout_times_start_end'], df_start_time)
else:
times_shootouts = []
# times_kills = np.round((np.array(gamedata_dict4player['times_kills']) - df_start_time) / TIMESTEP)
# times_deaths = np.round((np.array(gamedata_dict4player['times_is_killed']) - df_start_time) / TIMESTEP)
# times_shootouts = np.round((np.array(gamedata_dict4player['shootout_times_start_end']) - df_start_time) / TIMESTEP)
#
# # times_kills = [np.round(TIMESTEP * np.round(moment / TIMESTEP), 2) for moment in times_kills] # 2 here just in case.
# # times_deaths = [np.round(TIMESTEP * np.round(moment / TIMESTEP), 2) for moment in times_deaths] # 2 here just in case.
# # times_shootouts = [np.round(TIMESTEP * np.round(moment / TIMESTEP), 2) for moment in times_shootouts] # 2 here just in case.
df_resampled4player['step'] = timestamp2step(df_resampled4player['time'], df_start_time)
df_resampled4player.set_index('step', inplace=True)
df_resampled4player['kill'] = 0
df_resampled4player['death'] = 0
df_resampled4player['shootout'] = 0
n_steps = len(df_resampled4player)
for time_kill in times_kills:
if (0 <= time_kill < n_steps):
# df_resampled4player.loc[time_kill, 'kill'] = 1
df_resampled4player.loc[time_kill, 'kill'] += 1 # I fixed a bug
for time_death in times_deaths:
if (0 <= time_death < n_steps):
# df_resampled4player.loc[time_death, 'death'] = 1
df_resampled4player.loc[time_death, 'death'] += 1 # I fixed a bug
for time_shootout_start, time_shootout_end in times_shootouts:
assert time_shootout_start <= time_shootout_end
if (0 <= time_shootout_start < n_steps):
# df_resampled4player.loc[time_shootout_start:time_shootout_end, 'shootout'] = 1
df_resampled4player.loc[time_shootout_start:time_shootout_end+1, 'shootout'] += 1 # I fixed a bug
df_resampled4player['timedelta'] = pd.to_timedelta(df_resampled4player.index.values * TIMESTEP, unit='s')
df_resampled4player.reset_index(inplace=True)
df_resampled4player.drop(columns=['step', 'time'], inplace=True)
df_resampled4player.set_index('timedelta', inplace=True)
data_dict_resampled_merged[player_id] = df_resampled4player
joblib.dump(data_dict_resampled_merged, 'data/data_dict_resampled_merged')