-
Notifications
You must be signed in to change notification settings - Fork 3
/
get_oir_c2.py
256 lines (233 loc) · 11.1 KB
/
get_oir_c2.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
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sun Apr 22 16:26:42 2018
@author: weiss
4.22
1.增加了近月函数this_contract()
2.调整了部分语法结构适应合约key
3.调整了部分逻辑冗余
4.23
1.修复更新bug
4.30
1.修复近月合约数据缺失
5.7
1.改正近月合约定义
5.8
1.修正合并数据逻辑
5.10
1.修正合并数据部分的代码误删和错误
2.精简和优化部分代码结构
"""
import time as t
import datetime
import pandas as pd
import numpy as np
import os
import warnings
from WindPy import *
w.start()
warnings.filterwarnings("ignore")
class oir(object):
def __init__(self,homePath, updatebegin = 20100101, endDate = \
int(t.strftime('%Y%m%d',t.localtime(t.time()))) ,params = [5,10,20]):
self.homePath = homePath + '/'
self.tradeDateList = pd.read_csv(self.homePath +'tradeDateList.csv')
self.params = params
self.suffix = '.h5'
self.beginDate = updatebegin
self.time = t.time()
if t.localtime(self.time).tm_hour < 15:
self.workDate = int(t.strftime('%Y%m%d',\
t.localtime(self.time - 24 * 60 *60)))
else:
self.workDate = int(t.strftime('%Y%m%d',t.localtime(self.time)))
#确定所需更新的日期
if endDate < self.workDate:
self.workDate = endDate
def this_contract(self,windSymbol):
symbol = windSymbol.split('.')[0]
def change_spot(y_month):
weekday = datetime.strptime(y_month+'-01', "%Y-%m-%d").weekday()
if weekday <= 5:
return (14 + 5 - weekday)
else:
return (14 + 6)
def this_month(date):
day = np.int32(str(date)[6:8])
if day >= change_spot(str(date)[:4]+'-'+str(date)[4:6]):
month = np.int32(str(date)[4:6])
if month == 12:
return str(np.int32(str(date)[2:4])+1)+'01'
else:
return str(date)[2:4]+"%02d"%(month%12+1)
else:
return str(date)[2:6]
self.tradeDateList[symbol+'_contract'] = \
self.tradeDateList['tradeDate'].apply(lambda x : symbol+this_month(x))
self.tradeDateList.to_csv(self.homePath +'tradeDateList.csv', index=None)
def updateDataFromWind(self,windSymbol):
symbol = windSymbol.split('.')[0]
colNames = ['tradeDate','ranks','member_name','long_position',
'long_position_increase','short_position',
'short_position_increase','volume']
colNamesFinal = ['tradeDate','ranks','member_name','long_position',
'long_position_increase','short_position',
'short_position_increase','net_position',
'net_position_increase','volume','updatingTime']
colNamesCon = ['tradeDate','member_name','long_position',
'long_position_increase','short_position',
'short_position_increase','net_position',
'net_position_increase','volume','updatingTime']
#获取合约数据的函数
def getFutureoirByDate(beginDate,endDate,windSymbol,windCode,position):
data = w.wset("futureoir","startdate="+beginDate+";enddate="+
endDate+";varity="+windSymbol+";wind_code=" +
windCode + ";order_by=" + position +
";ranks=all;field=date,ranks,member_name,long_position,long_position_increase,short_position,short_position_increase,vol")
if len(data.Data) == 0:
return pd.DataFrame([])
dataout = pd.DataFrame()
try:
for i in range(len(colNames)):
dataout[colNames[i]] = data.Data[i]
except:
print(windSymbol + " cannot get data on " + date + ' !')
return pd.DataFrame([])
dataout['tradeDate'] = dataout['tradeDate'].astype(str)
dataout['tradeDate'] = pd.to_datetime(dataout['tradeDate'],\
format='%Y-%m-%d',errors='ignore')
dataout['net_position'] = dataout['long_position'] -\
dataout['short_position']
dataout['net_position_increase'] = \
dataout['long_position_increase'] \
- dataout['short_position_increase']
return dataout
dateList = pd.DataFrame()
dateList['tradeDate'] = self.tradeDateList['tradeDate'].astype(str)
dateList[symbol+'_contract'] = self.tradeDateList[symbol+'_contract']\
+'.'+ windSymbol.split('.')[1]
for position in ['long','short']:
endDate = str(self.workDate)
#如果存在数据，从上次更新日之后更新
status = 0
data = pd.DataFrame()
if os.path.exists(self.homePath + 'rank' + self.suffix):
try:
lastData = pd.read_hdf(self.homePath + 'rank' \
+ self.suffix, position +'_'+ windSymbol)
if len(lastData) == 0:
continue
lastDate = str(lastData['tradeDate'].iloc[-1])
lastDate = lastDate[0:4] + lastDate[5:7] + lastDate[8:10]
beginDate = dateList[dateList['tradeDate'] > lastDate]\
['tradeDate'].iloc[0]
beginDate = str(beginDate)
if beginDate > endDate:
continue
print(windSymbol+ '_' +position+ ', begin:' + beginDate +\
',end:' + endDate + ' updating...')
data = lastData
except:
status = 1
#不存在
else:
status = 1
if status == 1:
beginDate = str(self.beginDate)
print(windSymbol+ '_' +position+', begin:'+\
beginDate+' getting...')
tempDateList = dateList[dateList['tradeDate'] >= beginDate]
tempDateList = tempDateList[tempDateList['tradeDate'] <=\
endDate].reset_index(drop=True)
for i in range(len(tempDateList)):
date = tempDateList['tradeDate'][i]
contract = tempDateList[symbol+'_contract'][i]
print(date)
if data.empty:
data = getFutureoirByDate(date,date,windSymbol,\
contract,position)
else:
temdata = getFutureoirByDate(date,date,windSymbol,\
contract,position)
data = pd.concat([data,temdata])
data = data.reset_index(drop=True)
data['updatingTime'] = t.strftime('%Y-%m-%d %H:%M:%S')
data = data[colNamesFinal]
data.to_hdf(self.homePath + 'rank'+self.suffix, position + '_' +\
windSymbol)
def x_or_y(df):
c = df.columns
choise = np.sign((df[c[0]]-df[c[1]]).apply(np.sign)+1/2)
result = pd.DataFrame()
result[c[0][:-2]] = (df[c[0]]*(1+choise)+df[c[1]]*(1-choise))/2
if len(c)>2:
result[c[2][:-2]] = (df[c[2]]*(1+choise)+df[c[3]]*(1-choise))/2
return result
#生成连续数据
print('continous data merging...')
long_p = pd.read_hdf(self.homePath + 'rank'+self.suffix, \
'long_' + windSymbol)
short_p = pd.read_hdf(self.homePath + 'rank'+self.suffix, \
'short_' + windSymbol)
con_position = pd.merge(long_p.drop(['ranks','updatingTime'],axis = 1)\
.fillna(0),short_p.drop(['ranks','updatingTime'],\
axis = 1).fillna(0),on=['member_name','tradeDate'],\
how = 'outer').fillna(0)
con_position = con_position.sort_values(\
by=['tradeDate','long_position_x'],ascending = [True,False])
con_p = pd.DataFrame(data = [],\
index = range(len(con_position)),columns = colNamesCon)
con_position = con_position.reset_index()
for z in ['long_position','short_position','net_position']:
print(z +' merging...')
p_df = con_position[[z+'_x',z+'_y',z+'_increase_x',z+'_increase_y']]
con_p[[z,z+'_increase']] = x_or_y(p_df)
p_df = con_position[['volume_x','volume_y']]
print('volume merging...')
con_p['volume'] = x_or_y(p_df)
con_p['tradeDate'] = con_position['tradeDate']
con_p['member_name'] = con_position['member_name']
con_p['updatingTime'] = t.strftime('%Y-%m-%d %H:%M:%S')
con_p=con_p[colNamesCon]
con_p.to_hdf(self.homePath + 'rank'+self.suffix,windSymbol)
print (symbol + " futureoir source data update complete!")
return
def getSignal(self,windSymbol):
con_position = pd.read_hdf(self.homePath + 'rank'+self.suffix,windSymbol)
#强制默认参数为[5,10,20]，否则出错
sum_position = pd.DataFrame(data = [],index = range(len(con_position)),\
columns = ['tradeDate']+['long_position_increase_5']+\
['long_position_increase_10']+['long_position_increase_20']+\
['short_position_increase_5']+['short_position_increase_10']+\
['short_position_increase_20'])
#生成排名数据
j = 0
for i in range(len(con_position)):
if i == 0 or (con_position['tradeDate'][i] != \
con_position['tradeDate'][i-1]):
sum_position['tradeDate'][j] = con_position['tradeDate'][i]
for tem_i in range(len(self.params)):
sum_position['long_position_increase_'+str(self.params[tem_i])][j] = \
con_position['long_position_increase'][i+len(self.params)-1-tem_i]
sum_position['short_position_increase_'+str(self.params[tem_i])][j] = \
con_position['short_position_increase'][i+len(self.params)-1-tem_i]
j = j + 1
sum_position = sum_position.iloc[0:j]
#signal
signal = pd.DataFrame()
signal['tradeDate'] = sum_position['tradeDate']
for k in self.params:
signal['signal' + str(k)] = (sum_position['long_position_increase_'+str(k)].\
apply(np.sign) - sum_position['short_position_increase_'+str(k)].\
apply(np.sign))//2
print(windSymbol.split('.')[0] + ' signal complete !')
return signal
if __name__=='__main__':
homePath = '/Users/weiss/Downloads'
windSymbol = 'IF.CFE'
IF = oir(homePath,updatebegin = 20170101,endDate = 20180328)
IF.this_contract(windSymbol)
IF.updateDataFromWind(windSymbol)
sig = IF.getSignal(windSymbol)
sig.to_csv(homePath + '/signal.csv',index = None)