本文整理汇总了Python中DatabaseHandler.DatabaseHandler.import_network方法的典型用法代码示例。如果您正苦于以下问题：Python DatabaseHandler.import_network方法的具体用法？Python DatabaseHandler.import_network怎么用？Python DatabaseHandler.import_network使用的例子？那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类DatabaseHandler.DatabaseHandler的用法示例。
在下文中一共展示了DatabaseHandler.import_network方法的1个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: __init__
# 需要导入模块: from DatabaseHandler import DatabaseHandler [as 别名]
# 或者: from DatabaseHandler.DatabaseHandler import import_network [as 别名]
class Interface:
"""
This class is high level interface, which combining data obtaining and analytics.
"""
def __init__(self, con_path):
self.__db_handler = DatabaseHandler(con_path)
def run_update(self, teams_to_update, threads, use_proxy, proxy_file, update_opponents):
for team_id in teams_to_update:
spider_configuration = {'threads': threads,
'use_proxy': use_proxy,
'proxy_file': proxy_file,
'ignore_id': self.__db_handler.get_matches_id(team_id)}
adapter = DotabuffAdapter(self.__db_handler, spider_configuration)
adapter.update_team(team_id)
print('Team Updated :: {}'.format(team_id))
if update_opponents:
adapter.update_opponents(team_id)
print('Team Opponents Updated :: {}'.format(team_id))
def run_backtesting(self, dotalounge_hist_file, dotalounge_names_file, window):
dotalounge_hist = pd.DataFrame.from_csv(dotalounge_hist_file, header=None)
dotalounge_names = pd.DataFrame.from_csv(dotalounge_names_file, header=0, index_col=None)
names = dotalounge_names['names'].tolist()
index = dotalounge_names['id'].tolist()
delta = datetime.timedelta(days=window)
BANKS = {}
unique_dates = list(set(dotalounge_hist.ix[:, 6].tolist()))
i = 0
for date_str in unique_dates:
date = datetime.datetime.strptime(date_str, '%Y-%m-%d')
games = dotalounge_hist.loc[(dotalounge_hist.ix[:, 6] == date_str)]
dates = [date - delta, date]
analytics = DotaBetsAnalytics(self.__db_handler.import_network(dates=dates))
methods = [analytics.marginal_winrate,
analytics.joint_winrate,
analytics.neighbors_winrate]
for game in games.iterrows():
teams_names = [game[1][1], game[1][2]]
if teams_names[0] in names and teams_names[1] in names:
try:
teams_id = [index[names.index(teams_names[0])], index[names.index(teams_names[1])]]
result = [0, 0]
result[game[1][3]] = 1
coeff = [game[1][4], game[1][5]]
probabilities = [result[0]]
for method_ind in range(len(methods)):
method_winrate = methods[method_ind](teams_id)
probabilities += [method_winrate[0], method_winrate[1]]
evs = [coeff[0] * method_winrate[0] - 1, coeff[1] * method_winrate[1] - 1]
rand_fav = randint(0, 1)
rand_out = abs(rand_fav - 1)
rand_won = result[rand_fav] * coeff[rand_fav] - 1
vs_book_fav = 0 if coeff[0] >= coeff[1] else 1
vs_book_out = abs(vs_book_fav - 1)
vs_book_won = result[vs_book_fav] * coeff[vs_book_fav] - 1
best_fav = 0 if result[0] == 1 else 1
best_won = result[best_fav] * coeff[best_fav] - 1
book_fav = 0 if coeff[0] <= coeff[1] else 1
book_out = abs(book_fav - 1)
book_won = result[book_fav] * coeff[book_fav] - 1
my_fav = 0 if evs[0] >= evs[1] else 1
my_out = abs(my_fav - 1)
if evs[my_fav] >= 0 and evs[my_out] < 0:
my_won = result[my_fav] * coeff[my_fav] - 1
my_ev = evs[my_fav]
else:
my_won = 0
my_ev = 0
if method_ind not in BANKS:
BANKS[method_ind] = DataFrame(columns=('rand_won',
'vs_book_won',
'book_won',
'my_won'))
result_row = [rand_won, vs_book_won, book_won, my_won]
BANKS[method_ind].loc[len(BANKS[method_ind])] = result_row
except:
pass
i += 1
print(i, len(unique_dates))
for bank in BANKS:
BANKS[bank].cumsum(0).plot()
plt.show()
def run_analytics(self, teams_list, coeffs_list, window):
delta = datetime.timedelta(days=window)
#.........这里部分代码省略.........