-
Notifications
You must be signed in to change notification settings - Fork 106
/
celue_save.py
188 lines (169 loc) · 9.07 KB
/
celue_save.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
"""
为日线数据添加全部股票的历史策略买点列。
由于策略需要随时修改调整，因此单独写了策略写入文件，没有整合进readTDX_lday.py
"""
import os
import sys
import time
from multiprocessing import Pool, RLock, freeze_support
import numpy as np
import pandas as pd
from tqdm import tqdm
from rich import print
import CeLue # 个人策略文件，不分享
import func
import user_config as ucfg
# 变量定义
要剔除的通达信概念 = ["ST板块", ] # list类型。通达信软件中查看“概念板块”。
要剔除的通达信行业 = ["T1002", ] # list类型。记事本打开 通达信目录\incon.dat，查看#TDXNHY标签的行业代码。T1002=证券
def celue_save(file_list, HS300_信号, tqdm_position=None):
def lambda_update0(x):
if type(x) == float:
x = np.nan
elif x == '0.0':
x = np.nan
return x
# print('\nRun task (%s)' % os.getpid())
starttime_tick = time.time()
df_celue = pd.DataFrame()
if 'single' in sys.argv[1:]:
tq = tqdm(file_list)
else:
tq = tqdm(file_list, leave=False, position=tqdm_position)
for stockcode in tq:
tq.set_description(stockcode)
# process_info = f'[{(stocklist.index(stockcode) + 1):>4}/{str(len(stocklist))}] {stockcode}'
pklfile = ucfg.tdx['pickle'] + os.sep + stockcode + ".pkl"
df = pd.read_pickle(pklfile)
if 'del' in sys.argv[1:]:
if 'celue_buy' in df.columns:
del df['celue_buy']
if 'celue_sell' in df.columns:
del df['celue_sell']
df.set_index('date', drop=False, inplace=True) # 时间为索引。方便与另外复权的DF表对齐合并
if not {'celue_buy', 'celue_buy'}.issubset(df.columns):
df.insert(df.shape[1], 'celue_buy', np.nan) # 插入celue_buy列，赋值NaN
df.insert(df.shape[1], 'celue_sell', np.nan) # 插入celue_sell列，赋值NaN
else:
# 由于make_fq时fillna将最新的空的celue单元格也填充为0，所以先恢复nan
df['celue_buy'] = (df['celue_buy']
.apply(lambda x: lambda_update0(x))
.mask(df['celue_buy'] == 'False', False)
.mask(df['celue_buy'] == 'True', True)
)
df['celue_sell'] = (df['celue_sell']
.apply(lambda x: lambda_update0(x))
.mask(df['celue_sell'] == 'False', False)
.mask(df['celue_sell'] == 'True', True)
)
if True in df['celue_buy'].isna().to_list():
start_date = df.index[np.where(df['celue_buy'].isna())[0][0]]
end_date = df.index[-1]
celue2 = CeLue.策略2(df, HS300_信号, start_date=start_date, end_date=end_date)
celue_sell = CeLue.卖策略(df, celue2, start_date=start_date, end_date=end_date)
df.loc[start_date:end_date, 'celue_buy'] = celue2
df.loc[start_date:end_date, 'celue_sell'] = celue_sell
df.reset_index(drop=True, inplace=True)
df.to_csv(ucfg.tdx['csv_lday'] + os.sep + stockcode + '.csv', index=False, encoding='gbk')
df.to_pickle(ucfg.tdx['pickle'] + os.sep + stockcode + ".pkl")
lefttime_tick = int((time.time() - starttime_tick) / (file_list.index(stockcode) + 1)
* (len(file_list) - (file_list.index(stockcode) + 1)))
# 提取celue是true的列，单独保存到一个df，返回这个df
df_celue = df_celue.append(df.loc[df['celue_buy'] | df['celue_sell']])
# print(f'{process_info} 已用{(time.time() - starttime_tick):.2f}秒 剩余预计{lefttime_tick}秒')
df_celue['date'] = pd.to_datetime(df_celue['date'], format='%Y-%m-%d') # 转为时间格式
df_celue.set_index('date', drop=False, inplace=True) # 时间为索引。方便与另外复权的DF表对齐合并
return df_celue
if __name__ == '__main__':
print(f'附带命令行参数 del 完全重新生成策略信号, 参数 single 单进程执行(默认多进程)')
starttime = time.time()
df_hs300 = pd.read_csv(ucfg.tdx['csv_index'] + '/000300.csv', index_col=None, encoding='gbk', dtype={'code': str})
df_hs300['date'] = pd.to_datetime(df_hs300['date'], format='%Y-%m-%d') # 转为时间格式
df_hs300.set_index('date', drop=False, inplace=True) # 时间为索引。方便与另外复权的DF表对齐合并
HS300_信号 = CeLue.策略HS300(df_hs300)
stocklist = [i[:-4] for i in os.listdir(ucfg.tdx['pickle'])]
if 'del' in sys.argv[1:]:
print(f'检测到参数 del, 完全重新生成策略信号')
if 'single' in sys.argv[1:]:
print(f'检测到参数 single, 单进程执行')
df_celue = celue_save(stocklist, HS300_信号)
else:
# 多线程。好像没啥效果提升
# threads = []
# t_num = 4 # 线程数
# for i in range(0, t_num):
# div = int(len(stocklist) / t_num)
# mod = len(stocklist) % t_num
# if i+1 != t_num:
# # print(i, i * div, (i + 1) * div)
# threads.append(threading.Thread(target=celue_save, args=(stocklist[i*div:(i+1)*div], HS300_信号)))
# else:
# # print(i, i * div, (i + 1) * div + mod)
# threads.append(threading.Thread(target=celue_save, args=(stocklist[i*div:(i+1)*div+mod], HS300_信号)))
# # celue_save(stocklist, HS300_信号)
#
# print(threads)
# for t in threads:
# t.setDaemon(True)
# t.start()
#
# for t in threads:
# t.join()
# print("\n")
# 多进程
# print('Parent process %s' % os.getpid())
t_num = os.cpu_count() - 2 # 进程数 读取CPU逻辑处理器个数
freeze_support() # for Windows support
tqdm.set_lock(RLock()) # for managing output contention
p = Pool(processes=t_num, initializer=tqdm.set_lock, initargs=(tqdm.get_lock(),))
pool_result = [] # 存放pool池的返回对象列表
for i in range(0, t_num):
div = int(len(stocklist) / t_num)
mod = len(stocklist) % t_num
if i + 1 != t_num:
# print(i, i * div, (i + 1) * div)
pool_result.append(p.apply_async(celue_save, args=(stocklist[i * div:(i + 1) * div], HS300_信号, i)))
else:
# print(i, i * div, (i + 1) * div + mod)
pool_result.append(
p.apply_async(celue_save, args=(stocklist[i * div:(i + 1) * div + mod], HS300_信号, i)))
# celue_save(stocklist, HS300_信号)
# print('Waiting for all subprocesses done...')
p.close()
p.join()
# 处理celue汇总.csv文件。保存为csv文件，方便查看
df_celue = pd.DataFrame()
# 读取pool的返回对象列表。i.get()是读取方法。拼接每个子进程返回的df
for i in pool_result:
df_celue = df_celue.append(i.get())
# df_celue 是处理后的所有股票策略信号汇总文件。
# 下面处理自定义股票板块剔除
# 生成要剔除的股票列表 kicklist
print(f'生成股票列表, 共 {len(stocklist)} 只股票')
print(f'剔除通达信概念股票: {要剔除的通达信概念}')
kicklist = []
df = func.get_TDX_blockfilecontent("block_gn.dat")
# 获取df中blockname列的值是ST板块的行，对应code列的值，转换为list。用filter函数与stocklist过滤，得出不包括ST股票的对象，最后转为list
for i in 要剔除的通达信概念:
kicklist = kicklist + df.loc[df['blockname'] == i]['code'].tolist()
print(f'剔除通达信行业股票: {要剔除的通达信行业}')
df = pd.read_csv(ucfg.tdx['tdx_path'] + os.sep + 'T0002' + os.sep + 'hq_cache' + os.sep + "tdxhy.cfg",
sep='|', header=None, dtype='object')
for i in 要剔除的通达信行业:
kicklist = kicklist + df.loc[df[2] == i][1].tolist()
print("剔除科创板股票")
tdx_stocks = pd.read_csv(ucfg.tdx['tdx_path'] + '/T0002/hq_cache/infoharbor_ex.code',
sep='|', header=None, index_col=None, encoding='gbk', dtype={0: str})
kicklist = kicklist + tdx_stocks[0][tdx_stocks[0].apply(lambda x: x[0:2] == "68")].to_list()
stocklist = list(filter(lambda i: i not in kicklist, stocklist))
print(f'共 {len(stocklist)} 只候选股票')
# df_celue 剔除在kicklist中的股票
df_celue = df_celue[~df_celue['code'].isin(kicklist)]
print(f'保存独立"celue汇总.csv"文件')
df_celue = (df_celue
.drop(["open", "high", "low", "vol", "amount", "adj", "流通股", "流通市值", "换手率"], axis=1)
.sort_index()
.reset_index(drop=True)
)
df_celue.to_csv(ucfg.tdx['csv_gbbq'] + os.sep + 'celue汇总.csv', index=True, encoding='gbk')
print(f'用时 {(time.time() - starttime):.2f} 秒, 全部处理完成，程序退出')