def players(self):
allres = {}
for p in tqdm(self.plyrs[1:]):
ix, ming = 2 if self.RoP == 'regular' else 4, 500 if self.RoP == 'regular' else 50
if p[ix] and int(p[ix]) > ming:
pm = p[1].split('/')[-1][:-5]
# print(pm)
tmp, atds = self.singleplayer(pm)
if tmp:
allres[pm] = [tmp, atds]
writeToPickle('./data/playerRATs_%s.pickle' % self.RoP, allres)
def splitLineups(lines5):
print(list(lines5[0])[0], '...', list(lines5[0])[-1])
print(list(lines5[1])[0], '...', list(lines5[1])[-1])
lines_all = [[{}, {}], [{}, {}], [{}, {}],
[{}, {}]] # 0一人组合1二人组合2三人组合3四人组合 0常规赛1季后赛
for season in range(2002, 2021):
ss = '%d_%d' % (season, season + 1)
print(ss)
for i in range(2):
lines1234 = [{}, {}, {}, {}] # 单赛季按球队 0一人组合1二人组合2三人组合3四人组合
for tm in lines5[i][ss]:
# 球队lineups初始化
for ix in range(4):
if tm not in lines1234[ix]:
lines1234[ix][tm] = {}
lines_all[ix][i][ss] = {}
line5_tm = lines5[i][ss][tm] # 单队5人lineups
# print(tm, len(line5_tm))
for L in line5_tm:
pms = L.split(' ') # 遍历每组5人组合
# print(pms)
pms1234 = [[], [], [], []] # 0一人组合1二人组合2三人组合3四人组合
for c in range(1, 5): # 按照每组5人组合，使用（排列）组合方法穷尽所有1-4人的搭配
for iter in itertools.combinations(pms, c):
pms1234[c - 1].append(list(iter))
for ix, combs_i in enumerate(
pms1234): # 遍历1-4人的所有搭配，累加时间和正负值
for combs in combs_i:
combstr = ' '.join(sorted(combs))
if combstr not in lines1234[ix][tm]: # 每种组合初始化
lines1234[ix][tm][combstr] = [
MPTime('0:00.0'), 0
]
lines1234[ix][tm][combstr][0] += line5_tm[L][0]
lines1234[ix][tm][combstr][1] += line5_tm[L][1]
for ix in range(4):
for tm in lines1234[ix]:
lines_sorted = {}
for k in sorted(lines1234[ix][tm],
key=lines1234[ix][tm].__getitem__,
reverse=True): # 字典按值排序
lines_sorted[k] = lines1234[ix][tm][k]
lines_all[ix][i][ss][tm] = lines_sorted
for ix in range(4):
writeToPickle(
'D:/sunyiwu/stat/data/Lineups/anaSeason%dLineups.pickle' %
(ix + 1), lines_all[ix])
print(ave_score)
# [28317, 1933]
# [435372.0, 27686.0]
# [68541:22.2, 4634:40.1]
# ['0:9.4', '0:10.0']
# [538495.0, 33785.0]
# [1.2368618101301876, 1.2202918442534132]
for pm in plyrs:
for i in range(2):
if plyrs[pm][i][0]:
plyrs[pm][i][1] = plyrs[pm][i][1].average_acc(plyrs[pm][i][0])
# plyrs[pm][i][2] /= plyrs[pm][i][0]
plyrs[pm][i][3]['TOV'][0] /= plyrs[pm][i][0]
for ss in plyrs[pm][i][-1]:
if plyrs[pm][i][-1][ss][0]:
plyrs[pm][i][-1][ss][1] = plyrs[pm][i][-1][ss][1].average_acc(plyrs[pm][i][-1][ss][0])
# plyrs[pm][i][-1][ss][2] /= plyrs[pm][i][-1][ss][0]
plyrs[pm][i][-1][ss][3]['TOV'][0] /= plyrs[pm][i][-1][ss][0]
print('共%d名球员' % len(plyrs))
# for pm in plyrs:
# print(pm, plyrs[pm])
print(exchange_plays)
print(MSerror)
writeToPickle('./data/Enforce/player%sEnforce.pickle' % tartext[tar_item], plyrs)
writeToPickle('./data/Enforce/season%sEnforceRecord.pickle' % tartext[tar_item], [count_games_all, count_item_all, count_time_all,
average_time_all, count_score_all, average_score_all])
th = ss.find_all('th')[0].text
tds = ss.find_all('td')
if (th or tds[2].text == 'NBA') and\
not (th == 'Career' and tds[2].text == 'TOT') and\
tds[2].text == 'NBA':
if not th:
th = 'Career'
seasonAVE.append([th] + [x.text for x in tds])
# 写入csv文件
if seasonAVE:
if not os.path.exists('./data/players/%s' % pm):
os.mkdir('./data/players/%s' % pm)
if not os.path.exists('./data/players/%s/regularGames' % pm):
os.mkdir('./data/players/%s/regularGames' % pm)
df = pd.DataFrame(seasonAVE, columns=items)
writeToPickle(
'./data/players/%s/regularGames/seasonAVE.pickle' % pm, df)
df = pd.DataFrame(singleGAMES, columns=regular_items_en.keys())
df[df == ''] = np.nan
for col in df.columns:
df[col] = df[col].astype('category')
writeToPickle(
'./data/players/%s/regularGames/regularGameBasicStat.pickle' %
pm, df)
else:
print('球员未参加过NBA常规赛')
#%%
# -----季后赛-----
if last_season > 2018:
seasonAVE = []
singleGAMES = []
np.set_printoptions(suppress=True)
pd.options.display.expand_frame_repr = False
pd.options.display.width = 50
RoF = 0
gm = '201201070NJN'
game = Game(gm, 'playoff' if RoF else 'regular')
record = game.game_scanner()
# for i in record:
# print(i)
record = game.game_analyser(record)
record = game.game_analyser(record, T=1)
ss = gameMarkToSeason(gm)
rof = 'playoff' if RoF else 'regular'
season_dir = 'D:/sunyiwu/stat/data/seasons_scanned/%s/%s/' % (ss, rof)
writeToPickle(season_dir + gm + '_scanned.pickle', record)
for i in record:
print(i)
# print()
game.find_time_series(record)
game.start_of_quarter(record)
# 判断胜者
scores = [game.bxscr[0][x][0] for x in game.bxscr[0]]
print(list(game.bxscr[0]))
print(scores)
print('主队胜' if scores[0] < scores[1] else '客队胜')
rot = game.rotation(record)
for i in rot:
# 表头
items = [x.text for x in charts[1].find_all('th')]
items[0] = 'players'
chart.append(items)
for row in charts[2:]:
stats = row.find_all('td')
if len(stats) > 0:
if row.find('th').a:
pm = row.find('th').a.attrs['href'].split('/')[-1][:-5]
else:
if 'game' not in j:
continue
pm = row.find('th').text
chart.append([pm] + [x.text for x in stats])
cs.append(chart)
if cs:
boxscores.append(cs)
# 保存单场比赛数据
if regularOrPlayoff:
writeToPickle(seasonDir + '/playoff/%s_boxscores.pickle' % date, boxscores)
else:
writeToPickle(seasonDir + '/regular/%s_boxscores.pickle' % date, boxscores)
else:
ths = tr.find_all('th')
if len(ths) == 1 and ths[0].get_text().strip() == 'Playoffs':
# 找到季后赛分割线
print('switch to Playoffs')
regularOrPlayoff = 1
print('=' * 50)
if ivs_pm:
for pm in ivs_pm:
flag = 1
for r in game.gameflow[qtr]:
ind = 1 if r[1] else 5
if (len(r) > 2 and pm in r[ind]) or (
len(r) == 2 and 'Jump' in r[ind]
and pm in r[ind]):
flag = 0
# print(qtr, pm, r)
break
if flag:
if pm not in plyrs:
plyrs[pm] = [
[], [], [], [], []
] # 0gm, 1节次, 2+/-, 3胜负, 4是否比赛最后一节
diff = plus_minus(
rot[tick] if tick < len(rot) else
record[-1], rot[SoQ]['S'], tm)
plyrs[pm][0].append(gm)
plyrs[pm][1].append(qtr)
plyrs[pm][2].append(diff)
plyrs[pm][3].append(int(winner == tm))
plyrs[pm][4].append(
int(qtr == game.quarters - 1))
# print('ivs', pm, gm, qtr)
for pm in plyrs:
print(pm, plyrs[pm])
writeToPickle('./data/anaInvisiblePlayers.pickle', plyrs)
from klasses.stats_items import *
from tqdm import tqdm
import pandas as pd
# 计算球员赛季和生涯场均及总和并保存
pm2pn = LoadPickle('../data/playermark2playername.pickle')
for p in tqdm(list(pm2pn.keys())):
# print(p)
for RoP in ['regular', 'playoff']:
player = Player(p, RoP)
if player.exists and not isinstance(player.data, list):
res = []
for sss, ss in player.yieldSeasons():
sss = player.on_board_games(sss)
res += player.ave_and_sum(sss, type=2)
res[-2] = [ss] + res[-2]
res[-1] = [ss] + res[-1]
sss = player.on_board_games(player.data)
res += player.ave_and_sum(sss, type=2)
res[-2] = ['career'] + res[-2]
res[-1] = ['career'] + res[-1]
res = pd.DataFrame(res,
columns=['Season'] +
list(regular_items_en.keys() if RoP ==
'regular' else playoff_items_en.keys()))
# for col in res.columns:
# res[col] = res[col].astype('category')
writeToPickle(
'../data/players/%s/%sGames/%sSaCAaS.pickle' % (p, RoP, RoP),
res)
rs = i.find_all('a')
if rs != None:
for r in rs:
url, rn = r.attrs['href'], r.text
if rn not in refs:
refs[rn] = url
reftmp.append(rn)
else: # 裁判无个人URL
reftmp = i.text.split('\xa0')[1].split(', ')
elif 'Att' in i.text:
gtmp['A'] = int(i.text.split('\xa0')[1].replace(',', ''))
else:
gtmp['T'] = i.text.split('\xa0')[1]
if reftmp or gtmp:
res.append([gm[0], ', '.join(reftmp),
gtmp['A'] if 'A' in gtmp else -1,
gtmp['T'] if 'T' in gtmp else ''])
# print(time.time() - s)
if res:
df = pd.DataFrame(res, columns=['gm', 'Referees', 'Attendance', 'Time of Game'])
writeToPickle('./data/seasons_RAT/%s.pickle' % ss, df)
writeToPickle('./data/refereeURLs.pickle', refs)
for pm in tmplyrs[rh]:
if pm not in plyrs:
plyrs[pm] = {}
if ss not in plyrs[pm]:
# 0本队1对手 0reg1plf 0全部比赛数1客场数2主场数3stats 0客场1主场2全部 0第一节1第二节2上半场3第三节4第四节5下半场6全场
# 0FG 1FGA 2FG% 33P 43PA 53P% 6FT 7FTA 8FT% 9ORB 10DRB 11TRB 12AST 13STL 14BLK 15TOV 16PF 17PTS 18PACE
plyrs[pm][ss] = [[[0, 0, 0, 0, np.zeros((3, 7, 19))], [0, 0, 0, 0, np.zeros((3, 7, 19))]],
[[0, 0, 0, 0, np.zeros((3, 7, 19))], [0, 0, 0, 0, np.zeros((3, 7, 19))]]]
# 记录本队
plyrs[pm][ss][0][i][0] += 1
plyrs[pm][ss][0][i][rh + 1] += 1
plyrs[pm][ss][0][i][3] += gametime
plyrs[pm][ss][0][i][-1][rh, 6, :] += bx[rh]
plyrs[pm][ss][0][i][-1][2, 6, :] += bx[rh]
# 记录对手
op = 0 if rh else 1
plyrs[pm][ss][1][i][0] += 1
plyrs[pm][ss][1][i][op + 1] += 1
plyrs[pm][ss][0][i][3] += gametime
plyrs[pm][ss][1][i][-1][op, 6, :] += bx[op]
plyrs[pm][ss][1][i][-1][2, 6, :] += bx[op]
print(seasons[ss][0][0], list(seasons[ss][0][-1][2, 6, :] / seasons[ss][0][0] / 2))
print(seasons[ss][1][0], list(seasons[ss][1][-1][2, 6, :] / seasons[ss][1][0] / 2))
# if 'jamesle01' in plyrs:
# print(plyrs['jamesle01'][ss][0][0], list(plyrs['jamesle01'][ss][0][-1][2, 6, :] / plyrs['jamesle01'][ss][0][0]))
# print(plyrs['jamesle01'][ss][1][0], list(plyrs['jamesle01'][ss][1][-1][2, 6, :] / plyrs['jamesle01'][ss][1][0]))
writeToPickle('./data/leagueSeasonAverage.pickle', seasons)
writeToPickle('./data/playerSeasonAverage.pickle', plyrs)
gms = os.listdir(season_dir)
for gm in gms:
c = LoadPickle(season_dir + gm)
for q, qtr in enumerate(c):
for ix, r in enumerate(qtr):
if len(r) == 6 and 'enters' in (r[1] if r[1] else r[-1]):
ind = 1 if r[1] else 5
tmp = r[ind].split(' ')
pm1, pm2 = tmp[0], tmp[-1]
if pm1 == pm2:
print(gm, c[q][ix][ind])
url = 'https://www.basketball-reference.com/boxscores/pbp/%s.html' % gm[:-7]
plays = getCode(url, 'UTF-8')
plays = plays.find('table', class_='stats_table').find_all('tr')
for play in plays:
tdPlays = play.find_all('td')
if len(tdPlays) == 6:
for p in tdPlays:
if p.find_all('a'):
s = p.get_text().strip()
if 'enters' in s:
ps = s.split(' enters the game for ')
if len(ps) > 1 and ps[0] == ps[1]:
pms = []
for a in p.find_all('a'):
pms.append(a.attrs['href'].split('/')[-1].split('.')[0])
correct = '%s enters the game for %s' % (pms[0], pms[1])
c[q][ix][ind] = correct
print(c[q][ix][ind])
writeToPickle(season_dir + gm, c)
sentence_o = sentence_o + a.attrs[
'href'].split('/')[-1].rstrip(
'.html')
elif 'ejected from game' == sentence_o:
sentence_o = a.attrs['href'].split(
'/')[-1].rstrip(
'.html') + sentence_o
pTmp.append(sentence_o)
else:
pTmp.append(p.get_text().strip())
gameProcess[qtr].append(pTmp)
else:
pass
# 保存单场比赛数据
if regularOrPlayoff:
writeToPickle(seasonDir + '/playoff/%s.pickle' % date,
gameProcess)
else:
writeToPickle(seasonDir + '/regular/%s.pickle' % date,
gameProcess)
# 更新赛季比赛列表
if regularOrPlayoff:
seasonuPlayoffSmmary.append(gameDetails)
else:
seasonuRegularSmmary.append(gameDetails)
else:
ths = tr.find_all('th')
if len(ths) == 1 and ths[0].get_text().strip() == 'Playoffs':
# 找到季后赛分割线
print('switch to Playoffs')
regularOrPlayoff = 1
writeToPickle(seasonDir + '/seasonRegularSummary.pickle',
os.path.exists(seasonDir + '/regular/%s_shot.pickle' % date)):
# 比赛详细过程
if not tds[-4].a:
continue
gameURL = 'https://www.basketball-reference.com' + '/boxscores/shot-chart/' + tds[-4].a.attrs['href'].lstrip('/boxscores')
gamePage = getCode(gameURL, 'UTF-8')
charts = gamePage.find_all('div', class_='shot-area')
try:
assert len(charts) == 2
except:
print('%s：本场比赛缺失投篮点数据！' % date)
continue
shootings = [[], []]
for i in range(2):
shoots = charts[i].find_all('div')
for shoot in shoots:
shootings[i].append([shoot.attrs['style'], shoot.attrs['tip'], shoot.attrs['class']])
# 保存单场比赛数据
if regularOrPlayoff:
writeToPickle(seasonDir + '/playoff/%s_shot.pickle' % date, shootings)
else:
writeToPickle(seasonDir + '/regular/%s_shot.pickle' % date, shootings)
else:
ths = tr.find_all('th')
if len(ths) == 1 and ths[0].get_text().strip() == 'Playoffs':
# 找到季后赛分割线
print('switch to Playoffs')
regularOrPlayoff = 1
print('=' * 50)
def write2file(self):
if self.gameflow and self.file_name:
writeToPickle(self.file_name, self.gameflow)
print('写入文件%s成功' % self.file_name)
else:
messagebox.showinfo('提示', '没有内容！')
'Jalen Harris', '2021', '2021', 'G', '6-5', '195',
'August 14, 1998', 'Louisiana Tech, Nevada'
]
players['winstca01'] = [
'Cassius Winston', '2021', '2021', 'G', '6-1', '185', '',
''
]
players['beyty01'] = [
'Tyler Bey', '2021', '2021', 'F', '6-7', '215',
'February 10, 1998', ''
]
players['marshna01'] = [
'Naji Marshall', '2021', '2021', 'F', '6-7', '220',
'January 24, 1998', ''
]
players['okongon01'] = [
'Onyeka Okongwu', '2021', '2021', 'C', '6-8', '235',
'December 11, 2000', ''
]
players['haganas01'] = [
'Ashton Hagans', '2021', '2021', 'G', '6-3', '190',
'July 8, 1999', ''
]
writeToPickle('./data/playerBasicInformation.pickle', players)
#%%
pm2pn = {}
for k in list(players):
pm2pn[k] = players[k][0]
writeToPickle('./data/playermark2playername.pickle', pm2pn)
for y in range(cmps_shts_area_all.shape[0]):
for x in range(cmps_shts_area_all.shape[1]):
text = ax.text(x,
y,
'%.2f' % cmps_shts_area_all[cols[x]][y],
ha="center",
va="center",
color="w")
# h1 = plt.contourf(cmps_shts_area_all)
fig.colorbar(img, ax=ax)
plt.show()
writeToPickle('D:/sunyiwu/stat/data/winning_point.pickle',
[[cmps_all, cmps_shts_area_all, basic_all, count_game],
cmps_by_season])
# 5048 2016-2021
# [0.8098256735340729, ' eFG%']
# [0.803486529318542, ' biggest lead']
# [0.7977416798732171, ' biggest lead pts%']
# [0.7022583201267829, ' 3PT %']
# [0.6816561014263075, ' 2PT %']
# [0.6024167987321711, ' TOV%']
# [0.5919175911251982, ' pts off tovs']
# [0.5845879556259905, ' FT/FGA']
# [0.5711172741679873, ' ORB%']
# [0.5673534072900158, ' FTA/FGA']
# [0.5635895404120443, ' 3PT PTS%']
# [0.563391442155309, ' pts in the paint']
