class RaceData:
def __init__(self, start_date, end_date):
self.start_date = start_date
self.end_date = end_date
self.ext = LBExtract(start_date, end_date, False)
self.set_raceuma_df()
def set_raceuma_df(self):
self.raceuma_df = self.ext.get_raceuma_table_base()
def get_learning_df(self):
learning_df = self.raceuma_df[[
"競走コード", "馬番", "予想オッズ", "予想タイム指数", "デフォルト得点", "得点V1", "得点V2",
"得点V3", "投票直前単勝オッズ", "予想展開"
]]
return learning_df
def get_result_df(self):
result_df = self.raceuma_df[["競走コード", "馬番", "単勝配当"]]
return result_df
def _get_extract_object(self, start_date, end_date, mock_flag):
""" 利用するExtクラスを指定する """
print("-- check! this is LBLoad class: " +
sys._getframe().f_code.co_name)
ext = LBExtract(start_date, end_date, mock_flag)
return ext
def _get_extract_object(self, start_date, end_date, mock_flag):
""" 利用するExtクラスを指定する """
ext = LBExtract(start_date, end_date, mock_flag)
return ext
def __init__(self, start_date, end_date, mock_flag):
self.start_date = start_date
self.end_date = end_date
self.ext = LBExtract(start_date, end_date, mock_flag)
self._set_predata()
self.dbx = dropbox.Dropbox(mc.DROPBOX_KEY)
import my_config as mc
import pickle
## 各指数の適切な配分を計算する
## 勝ち指数：単勝回収率・勝率を重視
## 軸指数：複勝回収率・複勝率を重視
## 穴指数：１番人気との馬連の回収率・的中率を重視
# データ取得
start_date = '2019/01/01'
end_date = '2019/12/31'
mock_flag = False
ext = LBExtract(start_date, end_date, mock_flag)
sim = LBSimulation(start_date, end_date, mock_flag)
dict_path = mc.return_base_path(False)
intermediate_folder = dict_path + 'intermediate/'
def get_type_df_list(df, type):
df.rename(columns={"RACE_KEY": "競走コード", "UMABAN": "馬番", "predict_std": "予測値偏差"}, inplace=True)
win_df = df[df["target"] == "WIN_FLAG"][["競走コード", "馬番", "予測値偏差"]].rename(
columns={"予測値偏差": "偏差" + type})
jiku_df = df[df["target"] == "JIKU_FLAG"][["競走コード", "馬番", "予測値偏差"]].rename(
columns={"予測値偏差": "偏差" + type})
ana_df = df[df["target"] == "ANA_FLAG"][["競走コード", "馬番", "予測値偏差"]].rename(
columns={"予測値偏差": "偏差" + type})
return [win_df, jiku_df, ana_df]
from pulp import *
from modules.lb_extract import LBExtract
from modules.lb_transform import LBTransform
import numpy as np
import pandas as pd
import pickle
import my_config as mc
start_date = '2019/01/01'
end_date = '2019/12/31'
ext = LBExtract(start_date, end_date, False)
tr = LBTransform(start_date, end_date)
#ext.mock_flag = True
#ext.set_mock_path()
pd.set_option('display.max_columns', 200)
pd.set_option('display.max_rows', 200)
temp_df = ext.get_raceuma_table_base()
#temp_df = tr.normalize_raceuma_df(raceuma_base_df)
df = temp_df[["競走コード", "馬番", "デフォルト得点", "確定着順", "単勝配当", "複勝配当", "得点V3"]]
dict_path = mc.return_base_path(False)
intermediate_folder = dict_path + 'intermediate/'
with open(intermediate_folder + 'lb_v1_lb_v1/raceuma_ens/export_data.pkl',
'rb') as f:
lb_v1_df = pickle.load(f)
with open(intermediate_folder + 'lb_v2_lb_v2/raceuma_ens/export_data.pkl',
## create data
if run_mode == "init":
if test_flag:
start_date = '2020/4/01'
else:
start_date = '2019/12/01'
elif run_mode == "daily":
start_date = (dt.now().replace(day=1)).strftime('%Y/%m/%d')
else:
start_date = (dt.now() + timedelta(days=0)).strftime('%Y/%m/%d')
end_date = (dt.now() + timedelta(days=0)).strftime('%Y/%m/%d')
mock_flag = False
ext = LBExtract(start_date, end_date, mock_flag)
base_race_df = ext.get_race_table_for_view()[[
"データ区分", "競走コード", "月日", "距離", "トラック種別コード", "競走番号", "場名", "競走種別コード",
"競走条件名称", "トラックコード", "発走時刻", "頭数", "天候コード", "前３ハロン", "前４ハロン", "後３ハロン",
"後４ハロン", "馬場状態コード", "前半タイム", "予想勝ち指数", "ペース", "予想決着指数", "投票フラグ", "波乱度",
"ラップタイム", "競走名略称", "メインレース", "競走名", "UMAREN_ARE", "UMATAN_ARE",
"SANRENPUKU_ARE", "UMAREN_ARE_RATE", "UMATAN_ARE_RATE",
"SANRENPUKU_ARE_RATE"
]]
base_race_df.loc[:,
"年月"] = base_race_df["月日"].apply(lambda x: x.strftime("%Y%m"))
base_race_df.loc[:, "競走条件名称"] = base_race_df["競走条件名称"].replace(' ',
'').replace(
' ', '')
base_raceuma_df = ext.get_raceuma_table_base()[[
def __init__(self, start_date, end_date):
self.start_date = start_date
self.end_date = end_date
self.ext = LBExtract(start_date, end_date, False)
self.set_raceuma_df()