ai 加入eps_ttm_yoy

ai/training_data_builder.py

@@ -142,7 +142,7 @@ class TrainingDataBuilder(ABC):
         DataAll.reset_index(inplace=True)
         DataAll.ffill(inplace=True)
         DataAll['EPS_TTM_YOY'] = (DataAll['JIFU_SPX_OPEPS_CURRQ_TTM'] / DataAll['JIFU_SPX_OPEPS_CURRQ_TTM'].shift(
-            -252) - 1.0)
+            252) - 1.0)
         if (self._toForecast):
             # 处理CPI_YOY:美国城镇消费物价指数同比未经季 CPURNSA:美国消费者物价指数未经季调
             DataAllCopy = DataAll.copy()
@@ -166,30 +166,28 @@ class TrainingDataBuilder(ABC):
 
         # delete future information
         DataAll.drop(['futureR', 'yLabel'], axis=1, inplace=True)
-
         X = DataAll.iloc[:, 1:].values
 
         ###################
         # scale data
         labels = list(LABEL_RANGE.keys())
         scaler = MinMaxScaler(feature_range=(labels[-1], labels[0]))
-        # scaledX = scaler.fit_transform(X)
         DataScaler = scaler.fit(X)
         scaledX = DataScaler.transform(X)
         scaledX_forecast = None
-        if (self._toForecast):
+        if self._toForecast:
             scaledX_forecast = DataScaler.transform(X_forecast)
 
             X_train = scaledX
             y_train = y
             X_test = []
             y_test = []
-
+            date_index = []
         else:
             # Step 2: Split data into train set and test set
             X_train, X_test, y_train, y_test = train_test_split(scaledX, y, test_size=0.02, shuffle=False)
-
+            date_index = DataAll['date'][-len(X_test):-self._numForecastDays].to_numpy()
             # To avoid data leak, test set should start from numForecastDays later
-            X_test = X_test[self._numForecastDays:]
-            y_test = y_test[self._numForecastDays:]
-        return X_train, X_test, y_train, y_test, scaledX_forecast, forecastDay
+            X_test = X_test[:-self._numForecastDays]
+            y_test = y_test[:-self._numForecastDays]
+        return X_train, X_test, y_train, y_test, scaledX_forecast, forecastDay, date_index

basic/sp500.py

@@ -6,6 +6,7 @@ import pandas as pd
 import pytz
 import requests
 from openpyxl.reader.excel import load_workbook
+from py_jftech import sendmail
 
 
 def is_dst():
@@ -78,7 +79,7 @@ def list_files_sorted_by_name(directory, max_day=None):
 
 
 def fetch_sp500():
-    temp_file = Path(__file__).parent/'resources/sp-500.xlsx'
+    temp_file = Path(__file__).parent / 'resources/sp-500.xlsx'
     response = requests.get("https://www.spglobal.com/spdji/en/documents/additional-material/sp-500-eps-est.xlsx")
     # 确保请求成功
     if response.status_code == 200:
@@ -91,7 +92,7 @@ def fetch_sp500():
 
 def save_sp500():
     fetch_sp500()
-    files = list_files_sorted_by_name(Path(__file__).parent/'resources')[-2:]
+    files = list_files_sorted_by_name(Path(__file__).parent / 'resources')[-2:]
     compare_day = None
     for file in files:
         # 使用openpyxl加载Excel文件
@@ -103,13 +104,15 @@ def save_sp500():
             compare_day = report_day
         else:
             if compare_day != report_day:
-                wb.save(Path(__file__).parent/f'resources/sp-500-eps-est_USA{usa_close_day()}.xlsx')
+                new_path = Path(__file__).parent / f'resources/sp-500-eps-est_USA{usa_close_day()}.xlsx'
+                wb.save(new_path)
+                send_sp500('download sp500.', [new_path])
         # 关闭工作簿
         wb.close()
 
 
 def sync_sp500(day):
-    file = Path(__file__).parent/'resources/sp-500-eps-est_USA20241014.xlsx'
+    file = Path(__file__).parent / 'resources/sp-500-eps-est_USA20241014.xlsx'
     if day:
         files = list_files_sorted_by_name(Path(__file__).parent / 'resources', day)
         if files:
@@ -138,14 +141,23 @@ def sync_sp500(day):
         date_value = datetime.strptime(str(ws[f'A{i}'].value).split(' ')[0].strip(), '%m/%d/%Y') if type(
             ws[f'A{i}'].value) == str else ws[f'A{i}'].value
         if date_value < report_day:
+            # 日期只要是季度首日也设置red_date = red_release_date
             data = {'date': date_value,
                     'eps': ws[f'C{i}'].value}
             data["releaseDate"] = data['date'] + timedelta(days=1)
+            data["date"] = data['releaseDate']
             datas.append(data)
         elif date_value == get_quarter_end_date(report_day):
-            data = {'date': report_day,
-                    'eps': ws[f'C{i}'].value,
-                    'releaseDate': datetime.strptime(str(file)[-13:-5], "%Y%m%d")}
+            # 如果发布日是季度末，则red_date = red_release_date
+            if report_day == get_quarter_end_date(report_day):
+                data = {'date': date_value,
+                        'eps': ws[f'C{i}'].value}
+                data["releaseDate"] = data['date'] + timedelta(days=1)
+                data["date"] = data['releaseDate']
+            else:
+                data = {'date': report_day,
+                        'eps': ws[f'C{i}'].value,
+                        'releaseDate': datetime.strptime(str(file)[-13:-5], "%Y%m%d")}
             datas.append(data)
     for i in range(actuals_row + 1, ws.max_row):
         if ws[f'A{i}'].value is None:
@@ -154,6 +166,7 @@ def sync_sp500(day):
             ws[f'A{i}'].value) == str else ws[f'A{i}'].value,
                 'eps': ws[f'C{i}'].value}
         data["releaseDate"] = data['date'] + timedelta(days=1)
+        data["date"] = data['releaseDate']
         datas.append(data)
     wb.close()
     datas = pd.DataFrame(datas[::-1])
@@ -162,6 +175,12 @@ def sync_sp500(day):
     return datas.to_dict(orient="records")[-1::] if day else datas.to_dict(orient="records")
 
 
+def send_sp500(content, attach_paths):
+    receives = ['Tony.Wu.Home@gmail.com']
+    subject = 'sp500 eps download'
+    sendmail(receives=receives, copies=[], attach_paths=attach_paths, subject=subject, content=content)
+
+
 if __name__ == '__main__':
     # print(list_files_sorted_by_name(Path(__file__).parent / 'resources'))
     # save_sp500()