def initialize(self):
globalVars.initialize()
loadedDataList = load_material_data()
# TODO:use logger latter
print('We now have {} in our globalVar now'.format(loadedDataList))
try:
shiftedReturn = globalVars.materialData['pctChange'].get_shifted(
-1)
except AttributeError as ae:
print(ae)
raise AttributeError('There\'s no pctChange in globalVars')
except Exception as e:
print(e)
raise
shiftedReturn.metadata.update({'shiftN': -1})
shiftedReturn.name = 'shiftedReturn'
self.allTradeDatetime = shiftedReturn.timestamp
self.dependents.update({'shiftedReturn': shiftedReturn})
# TODO: should load from some factor.json file latter
############ this part realy socks
############ to be modified latter
############ with nice designed globalVars
############ the factor in globalVars.factors is a dict
toLoadFactors = ['close', 'high', 'low', 'open']
for aFactor in toLoadFactors:
simple_load_factor(aFactor)
self.factorNameList.append(aFactor)
def run(self):
# initialize the globalVars
globalVars.initialize()
self.logger.info('globalVars is initialized')
# load material data
loadedDataList = load_material_data()
self.logger.info('material data {0} is loaded'.format(loadedDataList))
# load factors
# TODO: should load from some factor.json file latter rather than simply load from material data
# self.factorNameList
# # 给globalVars注册factors（dict）
# # key：factor的名字，value：generalData
if 'factors' not in globalVars.varList:
globalVars.register('factors', {})
# TODO: factorManager
for factorName in self.factorNameList:
globalVars.factors[factorName] = Factor(
factorName, globalVars.materialData[factorName])
print(factorName, 'is now in globalVars.factors')
self.logger.info("factor {0} is loaded".format(factorName))
self.logger.info("all factors are loaded")
# calculate the signal
# 设置计算factors的参数
# 可直接传pipeline
# maskList：mask的nameList
self.logger.info("start to generate signalGenerator")
self.signalGenerator = self.signalGeneratorClass(
model=CrossSectionalModelXGBoost, logger=self.logger)
signals = self.signalGenerator.generate_signals(**self.params)
return signals
import random
from inspect import getmembers, isfunction
#%% initialize global vars
import numpy as np
from deap import base, creator, gp, tools
from Tool import globalVars
from Tool.GeneralData import GeneralData
from GetData.loadData import load_data
from GeneticPogramming import scalarFunctions, singleDataFunctions, singleDataNFunctions, coupleDataFunctions
#%% initialize global vars
globalVars.initialize()
loadedData = load_data()
globalVars.list_vars()
#%% add primitives
pset = gp.PrimitiveSetTyped('main', [
GeneralData, GeneralData, GeneralData, GeneralData, GeneralData,
GeneralData
], GeneralData)
pset.renameArguments(ARG0='CLOSE')
pset.renameArguments(ARG1='OPEN')
pset.renameArguments(ARG2='HIGH')
pset.renameArguments(ARG3='LOW')
pset.renameArguments(ARG4='AMOUNT')
pset.renameArguments(ARG5='VOLUME')
# 记录数据
record = stats.compile(pop)
logger.info("The {} th record:{}".format(gen, str(record)))
logbook.record(gen=gen, **record)
return (False, pop, logbook)
from GeneticPogramming.createMultiProcessWorker import toolbox, evaluate, materialDataNames, stats, logbook, compileFactor
from Tool import globalVars, Logger
from GetData import load_material_data, load_barra_data, align_all_to
# set up logger
loggerFolder = PROJECT_ROOT + "Tool\\log\\"
logger = Logger(loggerFolder, 'log')
globalVars.initialize(logger)
# load data to globalVars
load_material_data()
load_barra_data()
globalVars.logger.info('load all......done')
# prepare data
materialDataDict = {
k: globalVars.materialData[k]
for k in materialDataNames
} # only take the data specified in materialDataNames
barraDict = globalVars.barra
toRegFactorDict = {}
# get the return to compare
def run(self):
# initialize the globalVars
globalVars.initialize()
self.logger.info('globalVars is initialized')
# load material data
loadedDataList = load_material_data()
self.logger.info('material data {0} is loaded'.format(loadedDataList))
# load factors
# TODO: should load from some factor.json file latter rather than simply load from material data
# self.factorNameList
# toLoadFactors = ['close',
# 'high',
# 'low',
# 'open',
# 'volume'
# ]
# # 给globalVars注册factors（dict）
# # key：factor的名字，value：generalData
if 'factors' not in globalVars.varList:
globalVars.register('factors', {})
globalVars.factors['{}_factor'.format(self.toTestFactor.name)] = Factor('{}_factor'.format(self.toTestFactor.name),self.toTestFactor)
self.factorNameList.append('{}_factor'.format(self.toTestFactor.name))
# TODO: factorManager
# for factorName in toLoadFactors:
# globalVars.factors[factorName] = Factor(factorName, globalVars.materialData[factorName])
# print(factorName, 'is now in globalVars.factors')
# self.factorNameList.append(factorName)
# self.logger.info("factor {0} is loaded".format(factorName))
# self.logger.info("all factors are loaded")
# calculate the signal
# 设置计算factors的参数
# 可直接传pipeline
# maskList：mask的nameList
params = {
"startDate": pd.to_datetime('2020-01-01'),
"endDate": pd.to_datetime('2020-10-31'),
"panelSize": 3,
"trainTestGap": 1,
"maskList": None,
"deExtremeMethod": DeExtremeMethod.MeanStd(),
"imputeMethod": ImputeMethod.JustMask(),
"standardizeMethod": StandardizeMethod.StandardScaler(),
"pipeline": None,
"factorNameList": self.factorNameList,
# params for XGBoost
"modelParams": {
"jsonPath": None,
"paraDict": {
"n_estimators": 50,
"random_state": 42,
"max_depth": 2}
},
# metric function for machine learning models
"metric_func": mean_squared_error,
# smoothing params
"smoothing_params": None
}
self.logger.info("start to generate signalGenerator")
self.signalGenerator = self.signalGeneratorClass(model=CrossSectionalModelXGBoost, logger=self.logger)
signals = self.signalGenerator.generate_signals(**params)
return signals