-
Notifications
You must be signed in to change notification settings - Fork 2
/
collect_augment_data.py
472 lines (418 loc) · 19.1 KB
/
collect_augment_data.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
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
#collect data and sort them for multi-class classification
#I am going to try this first without cfbd data
from collect_data import read_api_key_from_yaml
import cfbd
import requests
from bs4 import BeautifulSoup
from pandas import DataFrame, read_csv, concat, to_numeric
import cfbd
from numpy import nan, where
from time import sleep
from os.path import join, exists
from os import getcwd, remove
from tqdm import tqdm
import yaml
"""
TODO: create a yaml file where for every year from 2024-2010, I have the top 30 teams
in the yaml files. This should help the models learn on patterns of top 30 teams
"""
def requests_sports_ref(URL):
while True:
try:
page = requests.get(URL)
soup = BeautifulSoup(page.content, "html.parser")
break
except:
print('HTTPSConnectionPool(host="www.sports-reference.com", port=443): Max retries exceeded. Retry in 10 seconds')
sleep(10)
return soup
def augment_team_names(team):
team = team.replace(' ', '-').lower()
if '.' in team:
team = team.replace(".", "")
if 'the' in team:
team = team.replace("the-", "")
if '&' in team:
team = team.replace("&", "")
if '(' in team and ')' in team:
team = team.replace("(", "")
team = team.replace(")", "")
if "'" in team:
team = team.replace("'", "")
if team == 'tcu':
team = 'texas-christian'
if team == 'lafayette':
team = 'louisiana-lafayette'
if team == 'louisiana':
team = 'louisiana-lafayette'
if team == 'utsa':
team = 'texas-san-antonio'
if team == 'utep':
team = 'texas-el-paso'
if team == 'sam-houston':
team = "sam-houston-state"
return team
def extract_features_from_other_team(team,team_playing,year):
sleep(2)
try:
URL = 'https://www.sports-reference.com/cfb/schools/' + team + '/' + str(year) + '/gamelog/'
soup = requests_sports_ref(URL)
table = soup.find(id="div_offense")
tbody = table.find('tbody')
except:
print(f'data for {team} does not exist')
return None
tr_body = tbody.find_all('tr')
game_result = []
turnovers = []
pass_cmp = []
pass_att = []
pass_yds = []
pass_td = []
rush_att = []
rush_yds = []
rush_td = []
rush_yds_per_att = []
tot_plays = []
tot_yds_per_play = []
first_down_pass = []
first_down_rush = []
first_down_penalty = []
first_down = []
penalty = []
penalty_yds = []
fumbles_lost = []
pass_int = []
game_result = []
#gets all games
for trb in tr_body:
#gets invidual game data
for td in trb.find_all('td'):
if td.get('data-stat') == "opp_name":
if '*' in td.get_text():
text_data = td.get_text().replace('*','')
else:
text_data = td.get_text()
opp_team = augment_team_names(text_data)
# print('=====')
# print(opp_team)
# print(team_playing)
# print('=====')
# input()
# if opp_team == team_playing:
# if td.get('data-stat') == "game_result":
# result = td.get_text()
if td.get('data-stat') == "turnovers":
turn = td.get_text()
if td.get('data-stat') == "pass_cmp":
pass_c = td.get_text()
if td.get('data-stat') == "pass_att":
pass_a = td.get_text()
if td.get('data-stat') == "pass_cmp_pct":
pass_c_p = td.get_text()
if td.get('data-stat') == "pass_yds":
pass_y= td.get_text()
if td.get('data-stat') == "pass_td":
pass_t = td.get_text()
if td.get('data-stat') == "rush_att":
rush_a = td.get_text()
if td.get('data-stat') == "rush_yds":
rush_y = td.get_text()
if td.get('data-stat') == "rush_yds_per_att":
rush_y_p_a = td.get_text()
if td.get('data-stat') == "rush_td":
rush_t = td.get_text()
if td.get('data-stat') == "tot_plays":
tot_p = td.get_text()
if td.get('data-stat') == "tot_yds_per_play":
tot_y_p_a = td.get_text()
if td.get('data-stat') == "first_down_pass":
first_d_p = td.get_text()
if td.get('data-stat') == "first_down_rush":
first_d_r = td.get_text()
if td.get('data-stat') == "first_down_penalty":
first_d_p = td.get_text()
if td.get('data-stat') == "first_down":
first_d = td.get_text()
if td.get('data-stat') == "penalty":
pen = td.get_text()
if td.get('data-stat') == "penalty_yds":
pen_y = td.get_text()
if td.get('data-stat') == "pass_int":
pass_i = td.get_text()
if td.get('data-stat') == "fumbles_lost":
fum_lost = td.get_text()
if opp_team == team_playing:
# game_result.append(result)
turnovers.append(turn)
pass_cmp.append(pass_c)
pass_att.append(pass_a)
pass_att.append(pass_c_p)
pass_yds.append(pass_y)
pass_td.append(pass_t)
rush_att.append(rush_a)
rush_yds.append(rush_y)
rush_yds_per_att.append(rush_y_p_a)
rush_td.append(rush_t)
tot_plays.append(tot_p)
first_down_penalty.append(first_d_p)
tot_yds_per_play.append(tot_y_p_a)
first_down_pass.append(first_d_p)
first_down_rush.append(first_d_r)
first_down.append(first_d)
penalty.append(pen)
penalty_yds.append(pen_y)
pass_int.append(pass_i)
fumbles_lost.append(fum_lost)
return DataFrame(list(zip(turnovers,pass_cmp,pass_att,pass_yds,
pass_td,
rush_att,
rush_yds,
rush_td,
rush_yds_per_att,
tot_plays,
tot_yds_per_play,
first_down_pass,
first_down_rush,
first_down_penalty,
first_down,
penalty,
penalty_yds,
fumbles_lost,
pass_int)),
columns =['turnovers_opp', 'pass_cmp_opp', 'pass_att_opp', 'pass_yds_opp', 'pass_td_opp', 'rush_att_opp',
'rush_yds_opp', 'rush_td_opp', 'rush_yds_per_att_opp', 'tot_plays_opp', 'tot_yds_per_play_opp',
'first_down_pass_opp', 'first_down_rush_opp', 'first_down_penalty_opp', 'first_down_opp', 'penalty_opp', 'penalty_yds_opp', 'fumbles_lost_opp',
'pass_int_opp'])
def collect_two_teams(URL,team,year):
soup = requests_sports_ref(URL)
table = soup.find(id="div_offense")
tbody = table.find('tbody')
tr_body = tbody.find_all('tr')
#current_team features
final_df = DataFrame()
for trb in tr_body:
game_result = []
turnovers = []
pass_cmp = []
pass_att = []
pass_yds = []
pass_td = []
rush_att = []
rush_yds = []
rush_td = []
rush_yds_per_att = []
tot_plays = []
tot_yds_per_play = []
first_down_pass = []
first_down_rush = []
first_down_penalty = []
first_down = []
penalty = []
penalty_yds = []
fumbles_lost = []
pass_int = []
game_loc = []
for td in trb.find_all('td'):
if td.get('data-stat') == "opp_name":
if '*' in td.get_text():
text_data = td.get_text().replace('*','')
else:
text_data = td.get_text()
opp_team = augment_team_names(text_data)
# print(f'opponent: {opp_team}')
opp_df = extract_features_from_other_team(opp_team,team,year)
#current team feature assignment
if td.get('data-stat') == "game_result":
# print(f'outcome: {td.get_text()}')
result = td.get_text()
if td.get('data-stat') == "turnovers":
turn = td.get_text()
if td.get('data-stat') == "pass_cmp":
pass_c = td.get_text()
if td.get('data-stat') == "pass_att":
pass_a = td.get_text()
if td.get('data-stat') == "pass_cmp_pct":
pass_c_p = td.get_text()
if td.get('data-stat') == "pass_yds":
pass_y= td.get_text()
if td.get('data-stat') == "pass_td":
pass_t = td.get_text()
if td.get('data-stat') == "rush_att":
rush_a = td.get_text()
if td.get('data-stat') == "rush_yds":
rush_y = td.get_text()
if td.get('data-stat') == "rush_yds_per_att":
rush_y_p_a = td.get_text()
if td.get('data-stat') == "rush_td":
rush_t = td.get_text()
if td.get('data-stat') == "tot_plays":
tot_p = td.get_text()
if td.get('data-stat') == "tot_yds_per_play":
tot_y_p_a = td.get_text()
if td.get('data-stat') == "first_down_pass":
first_d_p = td.get_text()
if td.get('data-stat') == "first_down_rush":
first_d_r = td.get_text()
if td.get('data-stat') == "first_down_penalty":
first_d_p = td.get_text()
if td.get('data-stat') == "first_down":
first_d = td.get_text()
if td.get('data-stat') == "penalty":
pen = td.get_text()
if td.get('data-stat') == "penalty_yds":
pen_y = td.get_text()
if td.get('data-stat') == "pass_int":
pass_i = td.get_text()
if td.get('data-stat') == "fumbles_lost":
fum_lost = td.get_text()
if td.get('data-stat') == "game_location":
if td.get_text() == '@':
loc = 0 #away
else:
loc = 1 #home
#team feature assignment
#if the data are not there for opp do not save
if opp_df is not None and isinstance(opp_df, DataFrame):
game_result.append(result)
turnovers.append(turn)
pass_cmp.append(pass_c)
pass_att.append(pass_a)
pass_att.append(pass_c_p)
pass_yds.append(pass_y)
pass_td.append(pass_t)
rush_att.append(rush_a)
rush_yds.append(rush_y)
rush_yds_per_att.append(rush_y_p_a)
rush_td.append(rush_t)
tot_plays.append(tot_p)
first_down_penalty.append(first_d_p)
tot_yds_per_play.append(tot_y_p_a)
first_down_pass.append(first_d_p)
first_down_rush.append(first_d_r)
first_down.append(first_d)
penalty.append(pen)
penalty_yds.append(pen_y)
pass_int.append(pass_i)
fumbles_lost.append(fum_lost)
game_loc.append(loc)
df = DataFrame(list(zip(game_result,turnovers,pass_cmp,pass_att,pass_yds,
pass_td,
rush_att,
rush_yds,
rush_td,
rush_yds_per_att,
tot_plays,
tot_yds_per_play,
first_down_pass,
first_down_rush,
first_down_penalty,
first_down,
penalty,
penalty_yds,
fumbles_lost,
pass_int,
game_loc)),
columns =['game_result','turnovers', 'pass_cmp', 'pass_att', 'pass_yds', 'pass_td', 'rush_att',
'rush_yds', 'rush_td', 'rush_yds_per_att', 'tot_plays', 'tot_yds_per_play',
'first_down_pass', 'first_down_rush', 'first_down_penalty', 'first_down', 'penalty', 'penalty_yds', 'fumbles_lost',
'pass_int','game_loc'])
df = concat([df, opp_df], axis=1)
final_df = concat([final_df, df])
sleep(4)
for column in final_df.columns:
if column != 'game_result':
final_df[column] = to_numeric(final_df[column], errors='coerce')
return final_df
def get_teams():
year_list_find = []
year_list = [2024,2023,2022,2021,2019,2018,2017,2016,2015,2014,2013,2012,2011,2010]#,2009,2008,2007,2006,2005,2004,2003,2002,2001,2000]
#all teams with data
all_team_train = True
# all_teams = get_teams_year(min(year_list),2023)
if all_team_train == True:
with open(join(getcwd(),'team_rankings_year.yaml')) as file:
data = yaml.load(file, Loader=yaml.FullLoader)
teams = data[2010]
teams_dict_year = {year: teams for year in range(2010, 2025)}
else:
#select only the top 40 teams from every year tested
with open(join(getcwd(),'team_rankings_year_top_40.yaml')) as file:
teams_dict_year = yaml.load(file, Loader=yaml.FullLoader)
# with open('top_30_teams.txt','r') as file:
# content = file.read()
# all_teams = content.split("\n")
# all_teams = [string for string in all_teams if string.strip() != ""]
if exists(join(getcwd(),'year_count.yaml')):
with open(join(getcwd(),'year_count.yaml')) as file:
year_counts = yaml.load(file, Loader=yaml.FullLoader)
#remove the current year to always update with the latest games
year_counts['year'].remove(2024)
else:
year_counts = {'year':year_list_find}
if year_counts['year']:
year_list_check = year_counts['year']
year_list_find = year_counts['year']
year_list = [i for i in year_list if i not in year_list_check]
print(f'Need data for year: {year_list}')
if year_list:
for year in year_list:
# all_teams = Teams(year)
# team_names = all_teams.dataframes.abbreviation
# team_names = team_names.sort_values()
final_list = []
x_feature_regress, y_feature_regress = [], []
all_teams = teams_dict_year[year]
for abv in tqdm(all_teams):
try:
print('')#tqdm thing
print(f'current team: {abv}, year: {year}')
# team = all_teams(abv)
str_combine = 'https://www.sports-reference.com/cfb/schools/' + abv.lower() + '/' + str(year) + '/gamelog/'
df_inst = collect_two_teams(str_combine,abv.lower(),year)
# print(df_inst)
final_list.append(df_inst)
if len(df_inst) % 2 != 0:
df_inst = df_inst.iloc[:-1]
x_feature_regress.append(df_inst.iloc[::2]) # Odd rows
y_feature_regress.append(df_inst.iloc[1::2]) # Even rows
except:
print(f'{abv} has no data for {year}')
output = concat(final_list)
final_data_x_regress = concat(x_feature_regress)
final_data_y_regress = concat(y_feature_regress)
final_data = output.replace(r'^\s*$', nan, regex=True) #replace empty string with NAN
if year == 2024:
#Remove old data so that you can update with the most recent games
if exists(join(getcwd(),'all_data_2024.csv')):
remove('all_data_2024.csv')
remove('x_feature_regression_2024.csv')
remove('y_feature_regression_2024.csv')
if not exists(join(getcwd(),'all_data_2024.csv')):
final_data.to_csv(join(getcwd(),'all_data_2024.csv'),index=False)
final_data_x_regress.to_csv(join(getcwd(),'x_feature_regression_2024.csv'),index=False)
final_data_y_regress.to_csv(join(getcwd(),'y_feature_regression_2024.csv'),index=False)
else:
if exists(join(getcwd(),'all_data.csv')):
all_data = read_csv(join(getcwd(),'all_data.csv'))
all_data = concat([all_data, final_data.dropna()])
all_data.to_csv(join(getcwd(),'all_data.csv'),index=False)
x_regress = read_csv(join(getcwd(),'x_feature_regression.csv'))
x_regress = concat([x_regress, final_data_x_regress])
x_regress.to_csv(join(getcwd(),'x_feature_regression.csv'),index=False)
y_regress = read_csv(join(getcwd(),'y_feature_regression.csv'))
y_regress = concat([y_regress, final_data_y_regress])
y_regress.to_csv(join(getcwd(),'y_feature_regression.csv'),index=False)
if not exists(join(getcwd(),'all_data.csv')):
final_data.to_csv(join(getcwd(),'all_data.csv'),index=False)
final_data_x_regress.to_csv(join(getcwd(),'x_feature_regression.csv'),index=False)
final_data_y_regress.to_csv(join(getcwd(),'y_feature_regression.csv'),index=False)
year_list_find.append(year)
print(f'year list after loop: {year_list_find}')
with open(join(getcwd(),'year_count.yaml'), 'w') as write_file:
yaml.dump(year_counts, write_file)
print(f'writing {year} to yaml file')
def main():
get_teams()
if __name__ == "__main__":
main()