修复一些bug，调整了目录结构

database/finhack_structure.sql

@@ -178,4 +178,4 @@ CREATE TABLE `factors_mining` (
 /*!40101 SET COLLATION_CONNECTION=@OLD_COLLATION_CONNECTION */;
 /*!40111 SET SQL_NOTES=@OLD_SQL_NOTES */;
 
--- Dump completed on 2023-12-27 13:30:47
+-- Dump completed on 2024-01-06 21:56:46

examples/demo-project/loader/testmodule_loader.py

@@ -7,7 +7,7 @@ def testaction(self):
         Log.logger.debug("loading "+self.module_name)
         Log.logger.debug("testarg1 is:"+str(self.args.testarg1))
         Log.logger.debug("testarg2 is:"+str(self.args.testarg2))
-        obj=self.klass()
+        obj=self.klass
         obj.args=self.args
         obj.run()
 

examples/demo-project/testmodule/default/default_testmodule.py

@@ -1,8 +1,10 @@
 import finhack.library.log as Log
 from runtime.constant import *
 import runtime.global_var as global_var
-
+from finhack.market.astock.astock import AStock
 import time
+from finhack.factor.default.factorManager import factorManager
+from finhack.factor.default.factorAnalyzer import factorAnalyzer
 class DefaultTestmodule():
     def __init__(self):
         pass
@@ -18,4 +20,10 @@ def run(self):
 
     def run2(self):
         print(self.args)
-        print('run2')
+        print('run2')
+        stock_list=AStock.getStockCodeList(strict=False, db='tushare')
+        print(stock_list)
+
+
+    def run3(self):
+        factorAnalyzer.alphalens("pe_0")

finhack/__init__.py

@@ -1 +1 @@
-__version__ = '0.0.1.dev4'
+__version__ = '0.0.1.dev5'

finhack/collector/tushare/tushare_collector.py

@@ -69,6 +69,9 @@ def run(self):
             table=list(v.values())[0]
             tsSHelper.setIndex(table,db)    
 
+
+    def save():
+        pass
 
 
     def getAStockBasic(self):

finhack/core/loader/collector_loader.py

@@ -5,7 +5,7 @@ def run(self):
         # print(self.module_path)
         # print(self.user_module_path)
         # print(self.klass)
-        collector=self.klass()
+        collector=self.klass
         collector.run()
 
         pass

finhack/core/loader/factor_loader.py

@@ -5,7 +5,7 @@ def run(self):
         # print(self.module_path)
         # print(self.user_module_path)
         # print(self.klass)
-        factor=self.klass()
+        factor=self.klass
         factor.run()
 
         pass

finhack/core/loader/helper_loader.py

@@ -5,7 +5,7 @@ def run(self):
         # print(self.module_path)
         # print(self.user_module_path)
         # print(self.klass)
-        helper=self.klass()
+        helper=self.klass
         helper.run()
 
         pass

finhack/core/loader/trader_loader.py

@@ -5,7 +5,7 @@ def run(self):
         # print(self.module_path)
         # print(self.user_module_path)
         # print(self.klass)
-        trader=self.klass()
+        trader=self.klass
         trader.args=self.args
         trader.run()
 

finhack/factor/default/alphaEngine.py

@@ -786,9 +786,6 @@ def calc(formula='',df=pd.DataFrame(),name="alpha",check=False,replace=False):
                 df=df.sort_index()
 
 
-
-
-
             if diff_date>0 and diff_date<100:
                 dt=datetime.datetime.strptime(str(max_date),'%Y%m%d')
                 start_date=dt-datetime.timedelta(days=700)

finhack/factor/default/default_factor.py

@@ -6,5 +6,12 @@
 from finhack.market.astock.astock import AStock
 from finhack.factor.default.taskRunner import taskRunner
 class DefaultFactor:
+    def __init__(self):
+        pass
+
     def run(self):
-        taskRunner.runTask()
+        pass
+        taskRunner.runTask()
+
+    def test(self):
+        print(self.args)

finhack/factor/default/factorAnalyzer.py

@@ -3,16 +3,133 @@
 import traceback
 import numpy as np
 import pandas as pd
-
+import pandas as pd
+import alphalens as al
+from alphalens.utils import get_clean_factor_and_forward_returns
+from alphalens.tears import create_full_tear_sheet
+import warnings
+warnings.simplefilter(action='ignore', category=FutureWarning)
+warnings.simplefilter(action='ignore', category=RuntimeWarning)
+warnings.simplefilter(action='ignore', category=UserWarning)
 from finhack.library.mydb import mydb
 from finhack.factor.default.factorManager import factorManager
+from finhack.market.astock.astock import AStock
+from scipy.stats import zscore
 
 class factorAnalyzer():
 
 
     def all_corr():
         pass
 
+
+
+    def alphalens(factor_name):
+
+
+        df_industry=AStock.getStockIndustry()
+
+
+        # df_all.index=df_all['date']
+        # price.index = pd.to_datetime(price.index)
+        # assets = df_all.set_index( [df_all.index,df_all['symbol']], drop=True,append=False, inplace=False)
+        df=factorManager.getFactors(factor_list=['close',factor_name])
+        # 假设 df 是您提供的 DataFrame，我们首先重置索引
+        df = df.reset_index().merge(df_industry, on='ts_code')
+        df['industry'] = df['industry'].fillna('其他')
+
+        df['trade_date'] = pd.to_datetime(df['trade_date'], format='%Y%m%d')
+
+        df[factor_name] = df.groupby(['trade_date', 'industry'])[factor_name].transform(zscore)
+
+        # 确保因子值没有 NaN，如果有 NaN，可以选择填充或者去除对应的行
+        df = df.dropna(subset=[factor_name,'industry'])
+
+        # 重置索引，准备进行 Alphalens 分析
+        df = df.set_index(['trade_date', 'ts_code'])
+
+        # 创建价格 DataFrame
+        prices = df['close'].unstack()
+
+        # 获取行业中性化后的因子数据
+        factor = df[factor_name]
+
+
+        unique_industries = df['industry'].unique()
+        # 创建 groupby_labels 字典，将每个行业标签映射到自己，确保没有遗漏
+        groupby_labels = {ind: ind for ind in unique_industries}
+
+
+        # 检查 groupby_labels 是否包含所有 unique_industries 中的行业
+        missing_labels = [ind for ind in unique_industries if ind not in groupby_labels]
+        if missing_labels:
+            print(f"Missing industry labels in groupby_labels: {missing_labels}")
+            # 您可以选择添加缺失的行业标签到 groupby_labels 中
+            for missing in missing_labels:
+                groupby_labels[missing] = '其他'  # 或者将其映射到 '其他'
+
+        # 使用 Alphalens 进行因子分析
+        factor_data = al.utils.get_clean_factor_and_forward_returns(
+            factor=factor,
+            prices=prices,
+            periods=(1, 5, 10),
+            groupby=df['industry'],  # 指定行业分组
+            groupby_labels=groupby_labels,  # 指定行业标签
+        )
+
+
+        # 因子收益率分析
+        mean_return_by_qt, std_err_by_qt = al.performance.mean_return_by_quantile(factor_data)
+        #aal.plotting.plot_quantile_returns_bar(mean_return_by_qt)
+        #aplt.show()
+
+        # 因子信息比率
+        ic_by_day = al.performance.factor_information_coefficient(factor_data)
+        #al.plotting.plot_information_coefficient(ic_by_day)
+        #plt.show()
+
+        # 分位数平均收益率
+        quantile_returns = al.performance.mean_return_by_quantile(factor_data)[0].apply(al.utils.rate_of_return, axis=0, base_period='1D')
+        #al.plotting.plot_quantile_returns_violin(quantile_returns)
+        #plt.show()
+
+        # 分位数累积收益
+        #cumulative_returns_by_qt = al.performance.cumulative_returns_by_quantile(factor_data, period=1)
+        #al.plotting.plot_cumulative_returns_by_quantile(cumulative_returns_by_qt, period=1)
+        #plt.show()
+
+        # 分位数收益率的全面统计
+        #full_tear_sheet = al.tears.create_full_tear_sheet(factor_data, long_short=True, group_neutral=False, by_group=False)
+
+        # 因子自相关性分析
+        autocorrelation = al.performance.factor_rank_autocorrelation(factor_data)
+        #al.plotting.plot_autocorrelation(autocorrelation)
+        #plt.show()
+
+        # 因子收益率和分位数收益率的IC分析
+        mean_monthly_ic = al.performance.mean_information_coefficient(factor_data, by_time='M')
+        #al.plotting.plot_monthly_ic_heatmap(mean_monthly_ic)
+        #plt.show()
+
+        print("\nmean_return_by_qt")
+        print(mean_return_by_qt)
+        print("\nic_by_day")
+        print(ic_by_day)
+        print("\nquantile_returns")
+        print(quantile_returns)
+        print("\nautocorrelation")
+        print(autocorrelation)
+        print("\nmean_monthly_ic")
+        print(mean_monthly_ic)
+
+        # print('---')
+        # print(full_tear_sheet)
+
+        #al.plotting.plot_quantile_returns_bar(mean_return_by_qt)
+        pass
+
+
+
     def analys(factor_name,df=pd.DataFrame(),days=[1,2,3,5,8,13,21],pool='all',start_date='20000101',end_date='20100101',formula="",relace=False,table='factors_analysis'):
         try:
 

finhack/factor/default/factorManager.py

@@ -164,7 +164,7 @@ def getAlphaList(listname):
 
     def getIndicatorsList():
         return_fileds=[]
-        path = os.path.dirname(__file__)+"/indicators/"
+        path = INDICATORS_DIR
         for subfile in os.listdir(path):
             if not '__' in subfile:
                 indicators=subfile.split('.py')

finhack/factor/default/indicatorCompute.py

@@ -12,7 +12,7 @@
 from functools import lru_cache
 from importlib import import_module
 from multiprocessing import cpu_count 
-
+import importlib
 from runtime.constant import *
 from finhack.library.mydb import mydb
 from finhack.library.config import Config
@@ -66,7 +66,9 @@ def split_list_n_list(origin_list, n):
         for code_list in code_lists:
             with ProcessPoolExecutor(max_workers=n) as pool:
                 for ts_code in code_list:
-                    mytask=pool.submit(indicatorCompute.computeListByStock,ts_code,list_name,'',factor_list,c_list)
+                    #computeListByStock(ts_code,list_name='all',where='',factor_list=None,c_list=[],pure=True,check=True,df_price=pd.DataFrame(),db='tushare'):
+
+                    mytask=pool.submit(indicatorCompute.computeListByStock,ts_code,list_name,'',factor_list,c_list,false,false)
                     tasklist.append(mytask)
 
         wait(tasklist,return_when=ALL_COMPLETED)
@@ -79,7 +81,7 @@ def split_list_n_list(origin_list, n):
 
     #计算单支股票的一坨因子
     #pure=True时，只保留factor_list中的因子
-    def computeListByStock(ts_code,list_name='all',where='',factor_list=None,c_list=[],pure=True,check=False,df_price=pd.DataFrame(),db='tushare'):
+    def computeListByStock(ts_code,list_name='all',where='',factor_list=None,c_list=[],pure=True,check=True,df_price=pd.DataFrame(),db='tushare'):
         try:
             Log.logger.info('computeListByStock---'+ts_code)
 
@@ -344,7 +346,22 @@ def computeFactorByStock(ts_code,factor_name,df_price=pd.DataFrame(),where='',db
                 df_price=indicatorCompute.computeFactorByStock(ts_code,f,df_price,db)
 
         factor=factor_name.split('_')
-        module = getattr(import_module('finhack.factor.default.indicators.'+indicators), indicators)
+        #module = getattr(import_module('finhack.factor.default.indicators.'+indicators), indicators)
+
+
+
+        # 定义文件路径
+        file_path = INDICATORS_DIR+indicators+".py"
+
+        # 获取文件名和类名
+        module_name = indicators
+        class_name = indicators
+
+        # 加载模块
+        spec = importlib.util.spec_from_file_location(module_name, file_path)
+        module = importlib.util.module_from_spec(spec)
+        spec.loader.exec_module(module)
+
         func=getattr(module,func_name,lambda x,y:x)
 
         shift="0"
@@ -424,7 +441,7 @@ def computeFactorByStock(ts_code,factor_name,df_price=pd.DataFrame(),where='',db
     def getFactorInfo(factor_name):
         factor=factor_name.split('_')
         factor_filed=factor[0]
-        path = os.path.dirname(__file__)+"/indicators/"
+        path = INDICATORS_DIR
         for subfile in os.listdir(path):
             if not '__' in subfile:
                 indicators=subfile.split('.py')

finhack/factor/default/preCheck.py

@@ -56,7 +56,7 @@ def checkIndicatorsChange():
             factor=factor_name.split('_')
             factor_filed=factor[0]
             return_fileds=[]
-            path = os.path.dirname(__file__)+"/indicators/"
+            path = INDICATORS_DIR
             for subfile in os.listdir(path):
                 if find:
                     continue