forked from harvey1673/pyktrader
-
Notifications
You must be signed in to change notification settings - Fork 3
/
strat_dt_chanfilter.py
207 lines (197 loc) · 9.92 KB
/
strat_dt_chanfilter.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
#-*- coding:utf-8 -*-
from base import *
from misc import *
import data_handler as dh
import copy
from strategy import *
class DTChanSplitTrader(Strategy):
def __init__(self, name, underliers, volumes,
agent = None,
trade_unit = [],
ratios = [],
lookbacks=[],
daily_close = False,
email_notify = None,
open_period = [300, 1500, 2115],
chan_func = {'func_high': [data_handler.DONCH_H, data_handler.donch_h],
'high_name': 'DONCH_H',
'func_low': [data_handler.DONCH_L, data_handler.donch_l],
'low_name': 'DONCH_L',
'func_args': {'n': 10}},
min_rng = [0.00]):
Strategy.__init__(self, name, underliers, volumes, trade_unit, agent, email_notify)
func_args = chan_func['func_args']
self.channel = func_args['n']
self.data_func = [
('d', BaseObject(name = chan_func['low_name'] + str(self.channel), \
sfunc=fcustom(chan_func['func_low'][0], **func_args), \
rfunc=fcustom(chan_func['func_low'][1], **func_args))),\
('d', BaseObject(name = chan_func['high_name'] + str(self.channel), \
sfunc=fcustom(chan_func['func_high'][0], **func_args), \
rfunc=fcustom(chan_func['func_high'][1], **func_args))),\
]
self.lookbacks = lookbacks
numAssets = len(underliers)
self.ratios = [[0.5, 0.5]] * numAssets
if len(ratios) > 1:
self.ratios = ratios
elif len(ratios) == 1:
self.ratios = ratios * numAssets
if len(lookbacks) > 0:
self.lookbacks = lookbacks
else:
self.lookbacks = [0] * numAssets
self.cur_rng = [0.0] * numAssets
self.chan_high = [0.0] * numAssets
self.chan_low = [0.0] * numAssets
self.tday_open = [0.0] * numAssets
self.open_period = open_period
self.open_idx = [0] * numAssets
self.tick_base = [0.0] * numAssets
self.order_type = OPT_LIMIT_ORDER
self.daily_close_buffer = 3
self.close_tday = [False] * numAssets
if len(daily_close) > 1:
self.close_tday = daily_close
elif len(daily_close) == 1:
self.close_tday = daily_close * numAssets
self.num_tick = 1
self.min_rng = [0.0] * numAssets
if len(min_rng) > 1:
self.min_rng = min_rng
elif len(min_rng) == 1:
self.min_rng = min_rng * numAssets
def initialize(self):
self.load_state()
low_str = self.data_func[0][1].name
high_str = self.data_func[1][1].name
for idx, underlier in enumerate(self.underliers):
inst = underlier[0]
self.tick_base[idx] = self.agent.instruments[inst].tick_base
min_id = self.agent.instruments[inst].last_tick_id/1000
min_id = int(min_id/100)*60 + min_id % 100 - self.daily_close_buffer
self.last_min_id[idx] = int(min_id/60)*100 + min_id % 60
ddf = self.agent.day_data[inst]
mdf = self.agent.min_data[inst][1]
last_date = ddf.index[-1]
self.chan_high[idx] = ddf.ix[-1, high_str]
self.chan_low[idx] = ddf.ix[-1, low_str]
self.open_idx[idx] = 0
if last_date < mdf.index[-1].date():
last_min = mdf['min_id'][-1]
pid = 0
for i in range(1, len(self.open_period)):
if self.open_period[i] >= last_min:
break
else:
pid = i
df = mdf[(mdf.index.date <= last_date)|(mdf['min_id'] < self.open_period[pid])]
post_df = mdf[(mdf.index.date > last_date) & (mdf['min_id'] >= self.open_period[pid])]
self.open_idx[idx] = pid
self.tday_open[idx] = post_df['open'][0]
else:
df = mdf
self.tday_open[idx] = mdf['close'][-1]
self.recalc_rng(idx, df)
self.save_state()
return
def recalc_rng(self, idx, df):
mdf = copy.copy(df)
inst = self.underliers[idx][0]
mdf.loc[:,'min_idx'] = pd.Series(0, index = mdf.index)
for i in range(1, len(self.open_period)-1):
mdf.loc[(mdf['min_id']>= self.open_period[i]) & (mdf['min_id']<self.open_period[i+1]), 'min_idx'] = i
mdf.loc[:, 'date_idx'] = mdf.index.date
ddf = mdf.groupby([mdf['date_idx'], mdf['min_idx']]).apply(dh.ohlcsum).reset_index().set_index('datetime')
win = self.lookbacks[idx]
if win > 0:
self.cur_rng[idx] = max(max(ddf.ix[-win:,'high'])- min(ddf.ix[-win:,'close']), max(ddf.ix[-win:,'close']) - min(ddf.ix[-win:,'low']))
elif win == 0:
self.cur_rng[idx] = max(max(ddf.ix[-2:,'high'])- min(ddf.ix[-2:,'close']), max(ddf.ix[-2:,'close']) - min(ddf.ix[-2:,'low']))
self.cur_rng[idx] = max(self.cur_rng[idx] * 0.5, ddf.ix[-1,'high']-ddf.ix[-1,'close'],ddf.ix[-1,'close']-ddf.ix[-1,'low'])
else:
self.cur_rng[idx] = max(ddf.ix[-1,'high']- ddf.ix[-1,'low'], abs(ddf.ix[-1,'close'] - ddf.ix[-2,'close']))
def save_local_variables(self, file_writer):
for idx, underlier in enumerate(self.underliers):
inst = underlier[0]
row = ['CurrRange', str(inst), self.cur_rng[idx]]
file_writer.writerow(row)
return
def load_local_variables(self, row):
if row[0] == 'CurrRange':
inst = str(row[1])
idx = self.under2idx[inst]
if idx >= 0:
self.cur_rng[idx] = float(row[2])
return
def register_bar_freq(self):
for idx, under in enumerate(self.underliers):
inst = under[0]
self.agent.inst2strat[inst][self.name].append(1)
#self.logger.debug("stat = %s register bar event for inst=%s freq = 1" % (self.name, inst, ))
def on_bar(self, idx, freq):
inst = self.underliers[idx][0]
min_id = self.agent.cur_min[inst]['min_id']
curr_min = self.agent.instruments[inst].last_update/1000
pid = self.open_idx[idx]
if pid < len(self.open_period)-1:
if (self.open_period[pid+1] > min_id) and (self.open_period[pid+1] <= curr_min):
self.recalc_rng(idx, self.agent.min_data[inst][1])
self.tday_open[idx] = self.agent.instruments[inst].price
self.logger.info("Note: the new split open is set to %s for inst=%s for stat = %s" % (self.tday_open[idx], inst, self.name, ))
def on_tick(self, idx, ctick):
if len(self.submitted_trades[idx]) > 0:
return
inst = self.underliers[idx][0]
if self.open_idx[idx] == 0:
self.tday_open[idx] = self.agent.cur_day[inst]['open']
if (self.tday_open[idx] <= 0.0) or (self.cur_rng[idx] <= 0) or (self.curr_prices[idx] <= 0.001):
self.logger.warning("warning: open price =0.0 or range = 0.0 or curr_price=0 for inst=%s for stat = %s" % (inst, self.name))
return
min_id = self.agent.tick_id/1000.0
num_pos = len(self.positions[idx])
buysell = 0
if num_pos > 1:
self.logger.warning('something wrong with position management - submitted trade is empty but trade position is more than 1')
return
elif num_pos == 1:
buysell = self.positions[idx][0].direction
tick_base = self.tick_base[idx]
t_open = self.tday_open[idx]
rng = max(self.cur_rng[idx] * self.ratios[idx][0], t_open * self.min_rng[idx])
buy_trig = t_open + rng
sell_trig = t_open - rng
if (min_id >= self.last_min_id[idx]):
if (buysell!=0) and (self.close_tday[idx]):
msg = 'DT to close position before EOD for inst = %s, direction=%s, volume=%s, current tick_id = %s' \
% (inst, buysell, self.trade_unit[idx], min_id)
self.close_tradepos(idx, self.positions[idx][0], self.curr_prices[idx] - buysell * self.num_tick * tick_base)
self.status_notifier(msg)
self.save_state()
return
if ((self.curr_prices[idx] >= buy_trig) and (buysell <=0)) or ((self.curr_prices[idx] <= sell_trig) and (buysell >=0)):
save_status = False
if buysell!=0:
msg = 'DT to close position for inst = %s, open= %s, buy_trig=%s, sell_trig=%s, curr_price= %s, direction=%s, volume=%s' \
% (inst, t_open, buy_trig, sell_trig, self.curr_prices[idx], buysell, self.trade_unit[idx])
self.close_tradepos(idx, self.positions[idx][0], self.curr_prices[idx] - buysell * self.num_tick * tick_base)
self.status_notifier(msg)
save_status = True
if self.trade_unit[idx] <= 0:
self.save_state()
return
if (self.curr_prices[idx] >= buy_trig):
buysell = 1
else:
buysell = -1
if self.curr_prices[idx] >= max(buy_trig, self.chan_high[idx]) or self.curr_prices[idx] <= min(sell_trig, self.chan_low[idx]):
msg = 'DT to open position for inst = %s, open= %s, buy_trig=%s, sell_trig=%s, curr_price= %s, direction=%s, volume=%s' \
% (inst, t_open, buy_trig, sell_trig, self.curr_prices[idx], buysell, self.trade_unit[idx])
self.open_tradepos(idx, buysell, self.curr_prices[idx] + buysell * self.num_tick * tick_base)
self.status_notifier(msg)
save_status = True
if save_status:
self.save_state()
return
def update_trade_unit(self):
pass