190 lines
8.9 KiB
Python
190 lines
8.9 KiB
Python
# -*- encoding:utf-8 -*-
|
||
"""
|
||
多支交易对象进行择时操作封装模块，内部通过AbuPickTimeWorker进行
|
||
择时，包装完善前后工作，包括多进程下的进度显示，错误处理捕获，结果
|
||
处理等事务
|
||
"""
|
||
|
||
from __future__ import absolute_import
|
||
from __future__ import print_function
|
||
from __future__ import division
|
||
|
||
import logging
|
||
|
||
import numpy as np
|
||
import pandas as pd
|
||
from enum import Enum
|
||
|
||
from .ABuPickTimeWorker import AbuPickTimeWorker
|
||
from ..CoreBu.ABuEnvProcess import add_process_env_sig
|
||
from ..TradeBu import ABuTradeExecute
|
||
from ..TradeBu import ABuTradeProxy
|
||
from ..TradeBu.ABuKLManager import AbuKLManager
|
||
from ..UtilBu.ABuProgress import AbuMulPidProgress
|
||
|
||
__author__ = '阿布'
|
||
__weixin__ = 'abu_quant'
|
||
|
||
|
||
class EFitError(Enum):
|
||
"""
|
||
择时操作的错误码
|
||
"""
|
||
|
||
# 择时操作正常完成，且至少生成一个order
|
||
FIT_OK = 0
|
||
# 择时对象数据获取错误
|
||
NET_ERROR = 1
|
||
# 择时对象数据错误
|
||
DATE_ERROR = 2
|
||
# 择时操作正常完成，但没有生成一个order
|
||
NO_ORDER_GEN = 3
|
||
# 其它错误
|
||
OTHER_ERROR = 4
|
||
|
||
|
||
def _do_pick_time_work(capital, buy_factors, sell_factors, kl_pd, benchmark, draw=False,
|
||
show_info=False, show_pg=False):
|
||
"""
|
||
内部方法：包装AbuPickTimeWorker进行fit，分配错误码，通过trade_summary生成orders_pd，action_pd
|
||
:param capital: AbuCapital实例对象
|
||
:param buy_factors: 买入因子序列
|
||
:param sell_factors: 卖出因子序列
|
||
:param kl_pd: 金融时间序列
|
||
:param benchmark: 交易基准对象，AbuBenchmark实例对象
|
||
:param draw: 是否绘制在对应的金融时间序列上的交易行为
|
||
:param show_info: 是否显示在整个金融时间序列上的交易结果
|
||
:param show_pg: 是否择时内部启动进度条，适合单进程或者每个进程里只有一个symbol进行择时
|
||
:return:
|
||
"""
|
||
if kl_pd is None or kl_pd.shape[0] == 0:
|
||
return None, EFitError.NET_ERROR
|
||
|
||
pick_timer_worker = AbuPickTimeWorker(capital, kl_pd, benchmark, buy_factors, sell_factors)
|
||
if show_pg:
|
||
pick_timer_worker.enable_task_pg()
|
||
pick_timer_worker.fit()
|
||
|
||
if len(pick_timer_worker.orders) == 0:
|
||
# 择时金融时间序列拟合操作后，没有任何order生成
|
||
return None, EFitError.NO_ORDER_GEN
|
||
|
||
# 生成关键的orders_pd与action_pd
|
||
orders_pd, action_pd, _ = ABuTradeProxy.trade_summary(pick_timer_worker.orders, kl_pd, draw=draw,
|
||
show_info=show_info)
|
||
|
||
# 最后生成list是因为tuple无法修改导致之后不能灵活处理
|
||
return [orders_pd, action_pd], EFitError.FIT_OK
|
||
|
||
|
||
@add_process_env_sig
|
||
def do_symbols_with_same_factors(target_symbols, benchmark, buy_factors, sell_factors, capital,
|
||
apply_capital=True, kl_pd_manager=None,
|
||
show=False, back_target_symbols=None, func_factors=None, show_progress=True):
|
||
"""
|
||
输入为多个择时交易对象，以及相同的择时买入，卖出因子序列，对多个交易对象上实施相同的因子
|
||
:param target_symbols: 多个择时交易对象序列
|
||
:param benchmark: 交易基准对象，AbuBenchmark实例对象
|
||
:param buy_factors: 买入因子序列
|
||
:param sell_factors: 卖出因子序列
|
||
:param capital: AbuCapital实例对象
|
||
:param apply_capital: 是否进行资金对象的融合，多进程环境下将是False
|
||
:param kl_pd_manager: 金融时间序列管理对象，AbuKLManager实例
|
||
:param show: 是否显示每个交易对象的交易细节
|
||
:param back_target_symbols: 补位targetSymbols为了忽略网络问题及数据不足导致的问题
|
||
:param func_factors: funcFactors在内层解开factors dicts为了do_symbols_with_diff_factors
|
||
:param show_progress: 进度条显示，默认True
|
||
"""
|
||
if kl_pd_manager is None:
|
||
kl_pd_manager = AbuKLManager(benchmark, capital)
|
||
|
||
def _batch_symbols_with_same_factors(p_buy_factors, p_sell_factors):
|
||
r_orders_pd = None
|
||
r_action_pd = None
|
||
r_all_fit_symbols_cnt = 0
|
||
# 启动多进程进度显示AbuMulPidProgress
|
||
with AbuMulPidProgress(len(target_symbols), 'pick times complete', show_progress=show_progress) as progress:
|
||
for epoch, target_symbol in enumerate(target_symbols):
|
||
# 如果symbol只有一个就不show了，留给下面_do_pick_time_work中show_pg内部显示进度
|
||
if len(target_symbols) > 1:
|
||
# 如果要绘制交易细节就不要clear了
|
||
progress.show(epoch + 1, clear=not show)
|
||
|
||
if func_factors is not None and callable(func_factors):
|
||
# 针对do_symbols_with_diff_factors mul factors等情况嵌入可变因子
|
||
p_buy_factors, p_sell_factors = func_factors(target_symbol)
|
||
try:
|
||
kl_pd = kl_pd_manager.get_pick_time_kl_pd(target_symbol)
|
||
ret, fit_error = _do_pick_time_work(capital, p_buy_factors, p_sell_factors, kl_pd, benchmark,
|
||
draw=show, show_info=show,
|
||
show_pg=(len(target_symbols) == 1 and show_progress))
|
||
except Exception as e:
|
||
logging.exception(e)
|
||
continue
|
||
|
||
if ret is None and back_target_symbols is not None:
|
||
# 择时结果错误或者没有order生成的情况下，如果有补位序列，择从序列中pop出一个，进行补位
|
||
if fit_error is not None and fit_error == EFitError.NO_ORDER_GEN:
|
||
# 没有order生成的要统计进去
|
||
r_all_fit_symbols_cnt += 1
|
||
while True:
|
||
if len(back_target_symbols) <= 0:
|
||
break
|
||
# pop出来代替原先的target
|
||
target_symbol = back_target_symbols.pop()
|
||
kl_pd = kl_pd_manager.get_pick_time_kl_pd(target_symbol)
|
||
ret, fit_error = _do_pick_time_work(capital, p_buy_factors, p_sell_factors, kl_pd, benchmark,
|
||
draw=show, show_info=show)
|
||
if fit_error == EFitError.NO_ORDER_GEN:
|
||
r_all_fit_symbols_cnt += 1
|
||
if ret is not None:
|
||
break
|
||
if ret is None:
|
||
continue
|
||
r_all_fit_symbols_cnt += 1
|
||
# 连接每一个交易对象生成的orders_pd和action_pd
|
||
r_orders_pd = ret[0] if r_orders_pd is None else pd.concat([r_orders_pd, ret[0]])
|
||
r_action_pd = ret[1] if r_action_pd is None else pd.concat([r_action_pd, ret[1]])
|
||
return r_orders_pd, r_action_pd, r_all_fit_symbols_cnt
|
||
|
||
orders_pd, action_pd, all_fit_symbols_cnt = _batch_symbols_with_same_factors(buy_factors, sell_factors)
|
||
if orders_pd is not None and action_pd is not None:
|
||
|
||
# 要sort'Date', 'action'两项，不然之后的行apply_action_to_capital后有问题
|
||
# noinspection PyUnresolvedReferences
|
||
action_pd = action_pd.sort_values(['Date', 'action'])
|
||
action_pd.index = np.arange(0, action_pd.shape[0])
|
||
# noinspection PyUnresolvedReferences
|
||
orders_pd = orders_pd.sort_values(['buy_date'])
|
||
if apply_capital:
|
||
# 如果非多进程环境下开始融合资金对象
|
||
ABuTradeExecute.apply_action_to_capital(capital, action_pd, kl_pd_manager, show_progress=show_progress)
|
||
|
||
return orders_pd, action_pd, all_fit_symbols_cnt
|
||
|
||
|
||
def do_symbols_with_diff_factors(target_symbols, benchmark, factor_dict, capital, apply_capital=True,
|
||
kl_pd_manager=None,
|
||
show=False,
|
||
back_target_symbols=None):
|
||
"""
|
||
输入为多个择时交易对象，每个交易对象有属于自己的买入，卖出因子，
|
||
在factor_dict中通过对象唯一标识进行提取
|
||
"""
|
||
|
||
def _func_factors(target_symbol):
|
||
"""
|
||
定义do_symbols_with_same_factors中使用的对交易因子dict进行解包的方法
|
||
"""
|
||
sub_dict = factor_dict[target_symbol]
|
||
buy_factors = sub_dict['buy_factors']
|
||
sell_factors = sub_dict['sell_factors']
|
||
return buy_factors, sell_factors
|
||
|
||
# 通过funcFactors在内层解开factors dict
|
||
return do_symbols_with_same_factors(target_symbols, benchmark, None, None, capital, apply_capital=apply_capital,
|
||
kl_pd_manager=kl_pd_manager,
|
||
show=show,
|
||
back_target_symbols=back_target_symbols,
|
||
func_factors=_func_factors)
|