-
Notifications
You must be signed in to change notification settings - Fork 3
/
forecast.py
182 lines (133 loc) · 5.57 KB
/
forecast.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
#forecast top 20 players' stats for next season
import pandas as pd
from bs4 import BeautifulSoup
import re
import urllib.request
import numpy as np
import csv
from mvp_model import zscore_model, stats_of_interest
FORECAST_FILE = 'data/mvpForecast.csv'
field_df = pd.read_csv('data/mvpForecast.csv')
player_names = field_df['Player']
player_names = player_names.tolist()
print(field_df)
start_ind = 7 #PER
def getPlayerStats(name, advanced=True):
player_name = name.lower()
ln_fi = player_name.find(' ') + 1 # index of first initial of last name
first = player_name[:2]
last = player_name[ln_fi:ln_fi + 5]
url = "https://www.basketball-reference.com/players/" + player_name[ln_fi] + "/" + last + first + "01.html"
if(name=='Anthony Davis'):
url = "https://www.basketball-reference.com/players/d/davisan02.html"
elif(name=='Clint Capela'):
url = "https://www.basketball-reference.com/players/c/capelca01.html"
elif(name=='D\'Angelo Russell'):
url = "https://www.basketball-reference.com/players/r/russeda01.html"
elif(name=='Kemba Walker'):
url = "https://www.basketball-reference.com/players/w/walkeke02.html"
print(url)
with urllib.request.urlopen(url) as response:
# UTF-8 doesn't support some initial character on the websites for some reason!
r = response.read().decode('latin-1')
content = re.sub(r'(?m)^\<!--.*\n?', '', r)
content = re.sub(r'(?m)^\-->.*\n?', '', content)
soup = BeautifulSoup(content, 'html.parser')
tables = soup.findAll('table')
if advanced:
table = tables[4]
else:
table = tables[0]
df = pd.read_html(str(table))[0]
return df
def getLatest(df):
latest_stats = df.iloc[-2].tolist()
age = latest_stats[df.columns.get_loc('Age')]+1
if len(df['VORP']) > 2:
latest_stats = np.array(list(filter(lambda a: str(a) != 'nan', df.iloc[-2].tolist()[start_ind:] )))
next_latest_stats = np.array(list(filter(lambda a: str(a) != 'nan', df.iloc[-3].tolist()[start_ind:] )))
return [latest_stats, next_latest_stats, age]
else:
latest_stats = np.array(list(filter(lambda a: str(a) != 'nan', df.iloc[-2].tolist()[start_ind:] )))
return [latest_stats, age]
def getDelta(stats_tuple):
mip_candidate = False
if len(stats_tuple) > 2 :
#progress cap for massive improvement players (e.g. Oladipo)
delta_PER = float(stats_tuple[0][0]-stats_tuple[1][0])
proportion = delta_PER/float(stats_tuple[0][0])
if proportion > 0.2:
mip_candidate = True
return [stats_tuple[0]-stats_tuple[1], stats_tuple[2], True, mip_candidate]
else:
return [stats_tuple[0], stats_tuple[1], False, mip_candidate]
def getChange(delta_tuple):
age = delta_tuple[1]
stat_delta = delta_tuple[0]
if delta_tuple[2] == True:
weight = (28-age)*0.15
#mip candidate
if delta_tuple[3] == True:
weight = 0.075
return weight*stat_delta
# only one year of exp
else:
weight = (28-age)*0.015
return weight*stat_delta
def updatedStatRow(df, change):
latest_stats = np.array(list(filter(lambda a: str(a) != 'nan', df.iloc[-2].tolist() )))
for i in range(len(latest_stats)-start_ind):
np.put(latest_stats, i+start_ind, float(latest_stats[i+start_ind])+float(change[i]))
return latest_stats
# ****************************************************************************
# EXPERIMENT **************************************
def adjustment(delta_tuple):
age = delta_tuple[1]
adjustment = 0.0
if age <= 28:
adjustment = (28 - age) * 0.025
else:
adjustment = (28 - age) * 0.01
weight = 1 + adjustment
return weight
def newUpdatedStatRow(df, adjustment):
latest_stats = np.array(list(filter(lambda a: str(a) != 'nan', df.iloc[-2].tolist())))
for i in range(len(latest_stats)-start_ind):
np.put(latest_stats, i+start_ind, adjustment*float(latest_stats[i+start_ind]))
return latest_stats
def newWriteCSV():
the_full_tensor_analytics_game_stats = []
header_row = field_df.columns.values.tolist()
the_full_tensor_analytics_game_stats.append(header_row)
for name in player_names:
df = getPlayerStats(name)
latest_szns = getLatest(df)
delta_tuple = getDelta(latest_szns)
change = adjustment(delta_tuple)
new_stat_row = newUpdatedStatRow(df, change).tolist()
new_stat_row.insert(0, name)
the_full_tensor_analytics_game_stats.append(new_stat_row)
print(the_full_tensor_analytics_game_stats)
with open(FORECAST_FILE, 'w', newline="") as f:
writer = csv.writer(f)
writer.writerows(the_full_tensor_analytics_game_stats)
# EXPERIMENT **************************************
# ****************************************************************************
def writeCSV():
the_full_tensor_analytics_game_stats = []
header_row = field_df.columns.values.tolist()
the_full_tensor_analytics_game_stats.append(header_row)
for name in player_names:
df = getPlayerStats(name)
latest_szns = getLatest(df)
delta_tuple = getDelta(latest_szns)
change = getChange(delta_tuple)
new_stat_row = updatedStatRow(df, change).tolist()
new_stat_row.insert(0, name)
the_full_tensor_analytics_game_stats.append(new_stat_row)
print(the_full_tensor_analytics_game_stats)
with open(FORECAST_FILE, 'w', newline="") as f:
writer = csv.writer(f)
writer.writerows(the_full_tensor_analytics_game_stats)
# writeCSV()
newWriteCSV()