siat 3.10.75__py3-none-any.whl → 3.10.125__py3-none-any.whl

siat/common.py

@@ -21,6 +21,27 @@ import pandas as pd
 SUFFIX_LIST_CN=['SS','SZ','BJ','SW','SH']
 SUFFIX_LIST_HK=['HK']
 #==============================================================================
+
+if __name__=='__main__':
+    ticker='000858.SZ'
+    ticker='AAPL'
+    
+    security_in_China(ticker)
+    
+def security_in_China(ticker):
+    """
+    功能：判断证券代码，是否在中国市场
+    """
+    
+    tlist=ticker.split('.')
+    if len(tlist) == 1: return False
+    else:
+        if tlist[1] in SUFFIX_LIST_CN:
+            return True
+        else:
+            return False
+    
+#==============================================================================    
 #设置全局语言环境
 import pickle
 
@@ -523,7 +544,7 @@ def portfolio_name(portfolio):
         name=portfolio[keylist[0]][2]
     except:
         #name="PF1"
-        e=text_lang("投资组合","Investment Portfolio")
+        name=text_lang("投资组合","Portfolio")
 
     return name    
 
@@ -912,6 +933,10 @@ def regparms(results):
     df_pvalues=pd.DataFrame(pvalues)
     df_pvalues.columns=['p_values']            
 
+    #取rsquared_adj值：单个数值，非数组
+    rsquared=results.rsquared
+    rsquared_adj=results.rsquared_adj
+    
     #生成星星
     df_pvalues['sig']=df_pvalues['p_values'].apply(lambda x:sigstars(x))
     
@@ -920,8 +945,10 @@ def regparms(results):
                     how='inner',left_index=True,right_index=True)
     parms2=pd.merge(parms1,df_pvalues, \
                     how='inner',left_index=True,right_index=True)
-
+    
+        
     return parms2
+
 #==============================================================================
 if __name__=='__main__':
     txt='QDII-指数'
@@ -2509,7 +2536,7 @@ def list2str(alist):
 
 #==============================================================================
 # FUNCTION TO REMOVE TIMEZONE
-def remove_timezone(dt):
+def remove_timezone_dt(dt):
    
     # HERE `dt` is a python datetime
     # object that used .replace() method
@@ -2517,7 +2544,7 @@ def remove_timezone(dt):
 #==============================================================================
 def remove_df_index_timezone(df):
     df['timestamp']=df.index
-    df['timestamp'] = df['timestamp'].apply(remove_timezone)
+    df['timestamp'] = df['timestamp'].apply(remove_timezone_dt)
     df.index=df['timestamp']
     del df['timestamp']
     
@@ -3760,6 +3787,7 @@ if __name__=='__main__':
     df_display_CSS(df,titletxt,footnote,facecolor,decimals)
 
 def df_display_CSS(df,titletxt='',footnote='',facecolor='papayawhip',decimals=2, \
+                   hide_columns=False,
                    first_col_align='left',second_col_align='right', \
                    last_col_align='right',other_col_align='right', \
                    titile_font_size='16px',heading_font_size='15px', \
@@ -3788,7 +3816,10 @@ def df_display_CSS(df,titletxt='',footnote='',facecolor='papayawhip',decimals=2,
     facecolor_default='papayawhip' 
     
     #不显示索引列，注意style1已经不是DaraFrame了
-    style1=df.style.hide() 
+    style1=df.style.hide()
+    
+    if hide_columns:
+        style1=df.style.hide(axis='index').hide(axis='columns')
     
     #设置数值字段的千分位符号，同时设置数值字段的小数点精度
     style2=style1.format(precision=decimals,thousands=',',na_rep='-') 
@@ -3966,10 +3997,21 @@ def df_index_timezone_remove(df):
     功能：去掉df索引日期中的时区信息，避免日期过滤时出错
     注意：从雅虎财经获取的数据中日期索引项很可能带有时区
     """
+    DEBUG=False
+    
+    # 检查是否因为处理时区而丢失了数据
+    if DEBUG:
+        print(f"BEFORE processing timezone, counts={len(df)}, from {df.index[0]} to {df.index[-1]}")
+    
     import pandas as pd
-    df.index = pd.to_datetime(df.index)
+    
+    #可能无法处理某些复杂的时区情况
+    df.index = pd.to_datetime(df.index,utc=True)
     df.index = df.index.tz_localize(None)
-
+    
+    if DEBUG:
+        print(f"AFTER processing timezone, counts={len(df)}, from {df.index[0]} to {df.index[-1]}")
+    
     return df
 #==============================================================================
 
@@ -5107,5 +5149,44 @@ def is_A_share(ticker):
         return False
     
     
+#==============================================================================
+
+def print2CSS(data_dict, \
+              titletxt='',footnote='',facecolor='papayawhip',decimals=2, \
+              hide_columns=True,
+              first_col_align='left',second_col_align='right', \
+              last_col_align='right',other_col_align='right', \
+              titile_font_size='14px',heading_font_size='14px', \
+              data_font_size='14px',footnote_font_size='11px'):
+    """
+    功能：将字典中的数据转化为df，使用CSS形式实现整齐输出
+    """
+    import pandas as pd
+    disp_df=pd.DataFrame(columns=['Item','Value'])    
+    
+    keys=list(data_dict.keys())
+    for key in keys:
+        value=data_dict[key]
+        
+        s=pd.Series({'Item':key,'Value':value}) 
+        disp_df=disp_df._append(s,ignore_index=True)        
+
+    df_display_CSS(disp_df,titletxt=titletxt,footnote=footnote, \
+                   facecolor=facecolor,decimals=decimals, \
+                       hide_columns=hide_columns,
+                   first_col_align=first_col_align,second_col_align=second_col_align, \
+                   last_col_align=last_col_align,other_col_align=other_col_align, \
+                   titile_font_size=titile_font_size,heading_font_size=heading_font_size, \
+                   data_font_size=data_font_size,footnote_font_size=footnote_font_size)
+        
+    return
+
+#==============================================================================
+#==============================================================================
 #==============================================================================
 #==============================================================================
+#==============================================================================
+#==============================================================================
+#==============================================================================
+#==============================================================================
+

siat/economy2.py

@@ -1367,6 +1367,68 @@ def economic_translate(indicator):
          'GDP per capita, PPP (constant 2021 international $)',
          'GDP per capita, PPP (constant 2021 international $)'],  
         
+        # NY.GNP.MKTP：国民总收入GNI总量=======================================
+        ['NY.GNP.MKTP.CD','GNI(美元现价)',
+         'GNI (current US$)',
+         'GNI (current US$)'],
+        
+        ['NY.GNP.MKTP.CN','GNI(本币现价)',
+         'GNI (current LCU)',
+         'GNI (current LCU)'],
+         
+        ['NY.GNP.MKTP.CN.AD','GNI(统计口径调整后，本币现价)',
+         'GNI: linked series (current LCU)',
+         'GNI: linked series (current LCU)'],       
+         
+        ['NY.GNP.MKTP.KD','GNI(2015美元不变价格)',
+         'GNI (constant 2015 US$)',
+         'GNI (constant 2015 US$)'],          
+         
+        ['NY.GNP.MKTP.KD.ZG','GNI年增速%(2015美元不变价格)',
+         'GNI growth (annual %, constant 2015 US$)',
+         'GNI growth (annual %, constant 2015 US$)'],          
+         
+        ['NY.GNP.MKTP.KN','GNI(本币不变价格)',
+         'GNI (constant LCU)',
+         'GNI (constant LCU)'],            
+         
+        ['NY.GNP.MKTP.PP.CD','GNI(购买力平价，国际美元现价)',
+         'GNI(PPP, current international $)',
+         'GNI(PPP, current international $)'],   
+         
+        ['NY.GNP.MKTP.PP.KD','GNI(购买力平价，2021国际美元不变价格)',
+         'GNI(PPP, constant 2021 international $)',
+         'GNI(PPP, constant 2021 international $)'],  
+        
+        # NY.GNP.PCAP：GNI人均=======================================
+        ['NY.GNP.PCAP.CD','人均GNI(美元现价)',
+         'GNI per capita (current US$)',
+         'GNI per capita (current US$)'],
+
+        ['NY.GNP.PCAP.CN','人均GNI(本币现价)',
+         'GNI per capita (current LCU)',
+         'GNI per capita (current LCU)'],        
+
+        ['NY.GNP.PCAP.KD','人均GNI(2015美元不变价格)',
+         'GNI per capita (constant 2015 US$)',
+         'GNI per capita (constant 2015 US$)'],   
+
+        ['NY.GNP.PCAP.KD.ZG','人均GNI年增速%(2015美元不变价格)',
+         'GNI per capita growth (annual %, constant 2015 US$)',
+         'GNI per capita growth (annual %, constant 2015 US$)'],  
+
+        ['NY.GNP.PCAP.KN','人均GNI(本币不变价格)',
+         'GNI per capita (constant LCU)',
+         'GNI per capita (constant LCU)'],  
+
+        ['NY.GNP.PCAP.PP.CD','人均GNI(购买力平价，国际美元现价)',
+         'GNI per capita, PPP (current international $)',
+         'GNI per capita, PPP (current international $)'],  
+
+        ['NY.GNP.PCAP.PP.KD','人均GNI(购买力平价，2021国际美元不变价格)',
+         'GNI per capita, PPP (constant 2021 international $)',
+         'GNI per capita, PPP (constant 2021 international $)'],  
+        
         #######################################################################
         #“International $”（国际美元）是一种标准化的货币单位，用于消除汇率差异，
         #使不同国家的经济指标（如 GDP）在购买力平价（PPP）基础上更具可比性。
@@ -1387,6 +1449,10 @@ def economic_translate(indicator):
         #######################################################################
         
         # GFDD.DM：证券市场-股票、权益与债券=====================================
+        ['CM.MKT.LCAP.GD.ZS','国内上市公司市值占GDP%',
+         'Market cap of domestic listed companies to GDP (%)',
+         'Market cap of domestic listed companies to GDP (%)'],        
+        
         ['GFDD.DM.01','股票市场总市值占GDP%',
          'Stock market capitalization to GDP (%)',
          'Stock market capitalization to GDP (%)'],
@@ -1494,7 +1560,15 @@ def economic_translate(indicator):
         ['FM.LBL.BMNY.CN','广义货币(本币现价)',
          'Broad money (current LCU)',
          'Broad money (current LCU)'],
-
+        
+        ['FM.LBL.BMNY.ZG','广义货币年增速%',
+         'Broad money growth (annual %)',
+         'Broad money growth (annual %)'],
+        
+        ['FM.LBL.MQMY.XD','货币流通速度(GDP/M2)',
+         'Income velocity of money (GDP/M2)',
+         'Income velocity of money (GDP/M2)'], 
+        
         ['FM.LBL.BMNY.GD.ZS','广义货币(占GDP%)',
          'Broad money (% of GDP)',
          'Broad money (% of GDP)'],

siat/fama_french.py

@@ -63,10 +63,112 @@ def convert2freq(df,new_freq='daily'):
     
     return df
 #==============================================================================
+#==============================================================================        
+if __name__=='__main__':
+    start='auto'; end='today'
+    
+    scope='US'
+    factor='FF3'
+    freq='yearly'
+    
+    show_ff_factors(ticker='US',indicator='FF3',freq='yearly')
 
+def show_ff_factors(market='US',indicator='FF3', \
+                    start='auto',end='today',freq='yearly'):
+    """
+    功能：套壳函数get_ff3_factors/get_ffc4_factors/get_ff5_factors
+    
+    【支持的因子种类(factor)】
+    FF3: FF3各个因子
+    FFC4: FFC4各个因子
+    FF5: FF5各个因子
+
+    【支持的国家/地区(scope)】
+    US: 美国
+    North_America：北美(美国+加拿大)
+    Global: 全球
+    Global_ex_US: 全球(除美国外)
+    Asia_Pacific_ex_Japan: 亚太(除日本外)
+    China: 中国
+    Japan: 日本
+    Europe: 欧洲
+
+    【支持的取样频率(freq)】
+    yearly: 年
+    monthly: 月
+    daily: 日
+    """
+    ticker=market
+    
+    indicator1=indicator.lower()
+    factorlist=['ff3','ffc4','ff5']
+    if not (indicator1 in factorlist):
+        print(f"  #Error(show_ff_factors): unsupported factor {indicator}")
+        print(f"  Supported factors for FF models: {factorlist}")
+        return
+    
+    import datetime; todaydt = datetime.date.today(); todaystr=str(todaydt)
+    if end == 'today': end=todaystr
+    
+    if start == 'auto':
+        if freq == 'yearly':
+            start=date_adjust(end,adjust=-365*5)
+        elif freq == 'monthly':
+            start=date_adjust(end,adjust=-31*8)
+        else: #freq=='daily'
+            start=date_adjust(end,adjust=-60)
+            
+    start,end=start_end_preprocess(start,end)
+    
+    freq=freq.lower()
+    if freq == 'daily':
+        print("Looking for daily factors, it may take time ...")
+    
+    if indicator1=='ff3':
+        df=get_ff3_factors(start=start,end=end,scope=ticker,freq=freq)
+    elif indicator1=='ffc4':
+        df=get_ffc4_factors(start=start,end=end,scope=ticker,freq=freq)
+    elif indicator1=='ff5':
+        df=get_ff5_factors(start=start,end=end,scope=ticker,freq=freq)
+    else:
+        pass
+    
+    #处理无数据情形
+    if df is None:
+        print(f"  #Warning(show_ff_factors): got none data from {start} to {end}")
+        print(f"  Suggestion: may specify an earlier start date and try again")
+        return
+    if len(df) == 0:
+        print(f"  #Warning(show_ff_factors): got zero data from {start} to {end}")
+        print(f"  Suggestion: may specify an earlier start date and try again")
+        return    
+    
+    #处理RF小数位数
+    df['RF']=df['RF'].apply(lambda x: str(round(x,4)))
+    
+    #titletxt="Asset Pricing Factors: "+indicator.upper()+'Model, '+freq+' @ '+ticker
+    titletxt=indicator.upper()+'模型定价因子：'+freq.title()+' @'+ticker
+    
+    if freq == 'yearly':
+        ft0="说明：RF为年化无风险利率%(美短债收益率)"
+    elif freq == 'monthly':
+        ft0="说明：RF为月度无风险利率%(美短债收益率)"
+    else:
+        ft0="说明：RF为日频无风险利率%(基于短期美债收益率)"
+        
+    #footnote="Data source: Dartmouth College, "+todaystr
+    footnote=ft0+'\n'+"数据来源：Dartmouth College，"+todaystr
+    
+    df_display_CSS(df,titletxt=titletxt,footnote=footnote,facecolor='papayawhip', \
+                   decimals=2, \
+                   first_col_align='center',second_col_align='center', \
+                   last_col_align='center',other_col_align='center')    
+
+    return
+
+#==============================================================================        
 if __name__=='__main__':
-    start='2025-1-1'
-    end='2025-6-13'
+    start='2020-1-1'; end='2025-4-30'
     
     scope='Europe'
     factor='Mom'
@@ -74,13 +176,14 @@ if __name__=='__main__':
         
     scope='US'
     factor='FF3'
+    freq='yearly'
     freq='daily'
     
     ff3d=get_ff_factors(start,end,scope,factor,freq='daily')
     ff3mth=get_ff_factors(start,end,scope,factor,freq='monthly')
     ff3y=get_ff_factors(start,end,scope,factor,freq='yearly')
 
-def get_ff_factors(start,end,scope='US',factor='FF3',freq='monthly'):
+def get_ff_factors(start,end,scope='US',factor='FF3',freq='monthly',retry=10):
     """
     功能：套壳函数get_ff_factors0，应对freq='daily'经常失败的变通方法
     """
@@ -88,7 +191,12 @@ def get_ff_factors(start,end,scope='US',factor='FF3',freq='monthly'):
     startpd=pd.to_datetime(start)
     endpd=pd.to_datetime(end)
     
-    fff=get_ff_factors0(start=start,end=end,scope=scope,factor=factor,freq=freq)
+    fff=None
+    for i in range(retry):
+        if not (fff is None): break
+        fff=get_ff_factors0(start=start,end=end,scope=scope,factor=factor,freq=freq)
+    
+    #多次尝试后仍未成功做变通处理：针对日度数据，使用月/年数据均匀化为日度数据
     if fff is None:
         if freq == 'daily':
             start1=date_adjust(start,adjust=-31)
@@ -117,6 +225,14 @@ def get_ff_factors(start,end,scope='US',factor='FF3',freq='monthly'):
     else:
         ff_factors=fff            
 
+    #加入日期
+    cols=list(ff_factors)
+    ff_factors['Date']=ff_factors.index
+    if freq == 'daily':
+        ff_factors['Date']=ff_factors['Date'].apply(lambda x: x.strftime("%Y-%m-%d"))
+    
+    ff_factors=ff_factors[['Date']+cols]
+    
     return ff_factors
 
 
@@ -812,7 +928,8 @@ def get_ffc4_factors(start,end,scope='US',freq='yearly'):
         Mom['date']=pd.to_datetime(Mom['date'].astype('str'))
         Mom.set_index('date',drop=True, inplace=True)
     
-    ffc4_factors=pd.merge(ff3,Mom,how='left',left_index=True,right_index=True)
+    #ffc4_factors=pd.merge(ff3,Mom,how='left',left_index=True,right_index=True)
+    ffc4_factors=pd.merge(ff3,Mom,how='inner',on='Date')
     #del ffc4_factors['date']
     
     return ffc4_factors
@@ -1003,11 +1120,15 @@ def reg_ff3_betas(ticker,start,end,scope='US',graph=True):
     ret=ret.rename(columns={'Close':'dRet'})
 
     #抓取FF3因子
+    print("  Looking for factors for FF3 model ...")
     freq='daily'
     ff3=get_ff3_factors(start,end,scope,freq)
     if ff3 is None:
         print("  #Warning(reg_ff3_betas): factors unavailable for",scope,freq,'from',start,'to',end)
         return None
+    if len(ff3) == 0:
+        print("  #Warning(reg_ff3_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
     ff3['Date']=ff3.index.strftime("%m/%d/%Y")
@@ -1032,7 +1153,7 @@ def reg_ff3_betas(ticker,start,end,scope='US',graph=True):
     #输出结果并绘制横向柱状图    
     if graph == True:
         gparms=parms.iloc[[1,2,3]]
-        print("\n",parms)
+        #print("\n",parms)
         title=ticker_name(ticker)+": FF3模型的贝塔系数"
         plt.title(title,fontsize=12,fontweight='bold')
         plt.ylabel("贝塔系数",fontsize=12,fontweight='bold')
@@ -1093,6 +1214,13 @@ if __name__=='__main__':
 
     
 #==============================================================================
+if __name__=='__main__':
+    ticker='MSFT'
+    start='2022-5-1'; end='2025-4-30'
+    scope='US'; graph=True
+    
+    
+    
 def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     """
     功能：测试一只股票对于FFC4各个因子的系数大小、符号方向和显著性
@@ -1121,16 +1249,20 @@ def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     ret=ret.rename(columns={'Close':'dRet'})
 
     #抓取每日FFC4因子
+    print("  Looking for factors for FFC4 model ...")
     freq='daily'
     ffc4=get_ffc4_factors(start,end,scope,freq)
     if ffc4 is None:
         print("  #Warning(reg_ffc4_betas): factors not available for",scope,freq,'from',start,'to',end)
         return None    
+    if len(ffc4) == 0:
+        print("  #Warning(reg_ffc4_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None    
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
-    ffc4['Date']=ffc4.index.strftime("%m/%d/%Y")
+    #ffc4['Date']=ffc4.index.strftime("%m/%d/%Y")
     ffc4['Date']=pd.to_datetime(ffc4['Date'])
-    ffc4.set_index('Date',inplace=True)    
+    ffc4.set_index('Date',drop=True,inplace=True)    
 
     #合成股票日收益率+每日FF3因子
     sample=pd.merge(ret,ffc4,how='inner',left_index=True,right_index=True)
@@ -1152,7 +1284,7 @@ def reg_ffc4_betas(ticker,start,end,scope='US',graph=True):
     #输出结果并绘制横向柱状图
     gparms=parms.iloc[[1,2,3,4]]
     if graph == True:
-        print("\n",parms)
+        #print("\n",parms)
         title=ticker_name(ticker)+": FFC4模型的贝塔系数"
         plt.title(title,fontsize=12,fontweight='bold')
         plt.ylabel("贝塔系数",fontsize=12,fontweight='bold')
@@ -1213,6 +1345,11 @@ if __name__=='__main__':
     test_ffc4_betas() 
     
 #==============================================================================
+if __name__=='__main__':
+    ticker='MSFT'
+    start='2024-5-1'; end='2025-4-30'
+    scope='US'; graph=True
+
 def reg_ff5_betas(ticker,start,end,scope='US',graph=True):
     """
     功能：测试一只股票对于FF5各个因子的系数大小、符号方向和显著性
@@ -1240,12 +1377,16 @@ def reg_ff5_betas(ticker,start,end,scope='US',graph=True):
     #命名日收益率字段为dRet
     ret=ret.rename(columns={'Close':'dRet'})
 
-    #抓取每日FF3因子
+    #抓取每日FF5因子
+    print("  Looking for factors for FF5 model ...")
     freq='daily'
     ff5=get_ff5_factors(start,end,scope,freq)
     if ff5 is None:
         print("  #Warning(reg_ff5_betas): factors not available for",scope,freq,'from',start,'to',end)
         return None       
+    if len(ff5) is None:
+        print("  #Warning(reg_ff5_betas): zero factors retrieved for",scope,freq,'from',start,'to',end)
+        return None       
     
     #改造索引类型为时间戳索引（DatetimeIndex），便于与ret合成
     ff5['Date']=ff5.index.strftime("%m/%d/%Y")
@@ -1269,7 +1410,7 @@ def reg_ff5_betas(ticker,start,end,scope='US',graph=True):
     #输出结果并绘制横向柱状图
     gparms=parms.iloc[[1,2,3,4,5]]
     if graph == True:
-        print("\n",parms)
+        #print("\n",parms)
         title=ticker_name(ticker)+"：FF5模型的贝塔系数"
         plt.title(title,fontsize=12,fontweight='bold')
         plt.ylabel("贝塔系数",fontsize=12,fontweight='bold')
@@ -1329,4 +1470,52 @@ def test_ff5_betas():
 if __name__=='__main__':
     test_ff5_betas() 
     
-#==============================================================================    
+#==============================================================================
+
+
+
+def reg_ff_betas(ticker,start,end='today',indicator='FF3',market='US'):
+    """
+    功能：套壳函数reg_ff3_betas/reg_ffc4_betas/reg_ff5_betas
+    参数：
+    indicator：默认'FF3',还可为'FFC4'或'FF5'
+    scope：默认'US',还可为'China'或'Japan'或'Europe'等
+    """
+    scope=market
+    
+    start,end=start_end_preprocess(start,end)
+    
+    indicator1=indicator.lower()
+    factorlist=['ff3','ffc4','ff5']
+    if not (indicator1 in factorlist):
+        print(f"  #Error(show_ff_factors): unsupported factor {indicator}")
+        print(f"  Supported factors for FF models: {factorlist}")
+        return    
+        
+    if indicator1=='ff3':
+        df=reg_ff3_betas(ticker=ticker,start=start,end=end,scope=scope)
+    elif indicator1=='ffc4':
+        df=reg_ffc4_betas(ticker=ticker,start=start,end=end,scope=scope)
+    elif indicator1=='ff5':
+        df=reg_ff5_betas(ticker=ticker,start=start,end=end,scope=scope)
+    else:
+        pass   
+    
+    #打印回归结果
+    cols=list(df)
+    df['const']=df.index
+    df=df[['const']+cols]
+    
+    titletxt=ticker_name(ticker)+"："+indicator.upper()+'模型回归结果'  
+    import datetime; todaydt = datetime.date.today(); todaystr=str(todaydt)
+    footnote="数据来源：Dartmouth College/sina/stooq/Yahoo等，"+todaystr    
+    
+    df.rename(columns={'const':'Constant','coef':'Coefficient', \
+                       'sig':'Significance'},inplace=True)
+    
+    df_display_CSS(df,titletxt=titletxt,footnote=footnote,facecolor='papayawhip', \
+                   decimals=4, \
+                   first_col_align='center',second_col_align='center', \
+                   last_col_align='center',other_col_align='center')       
+        
+    return df