analyser_hj3415 2.9.3__tar.gz → 2.9.5__tar.gz

{analyser_hj3415-2.9.3 → analyser_hj3415-2.9.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: analyser_hj3415
-Version: 2.9.3
+Version: 2.9.5
 Summary: Stock analyser and database processing programs
 Requires-Python: >=3.6
 Description-Content-Type: text/markdown
@@ -12,6 +12,7 @@ Requires-Dist: yfinance>=0.2.44
 Requires-Dist: prophet>=1.1.6
 Requires-Dist: kaleido>=0.2.1
 Requires-Dist: matplotlib>=3.9.2
+Requires-Dist: tensorflow>=2.18.0
 
 ### analyser-hj3415
 

{analyser_hj3415-2.9.3 → analyser_hj3415-2.9.5}/analyser_hj3415/tsa.py

@@ -1,6 +1,8 @@
 """
 Time Series Analysis
 """
+from pprint import pprint
+
 import numpy as np
 import yfinance as yf
 from datetime import datetime, timedelta
@@ -98,6 +100,8 @@ class MyProphet:
 
         # 추가 변수를 정규화
         df['volume_scaled'] = self.scaler.fit_transform(df[['volume']])
+        tsa_logger.debug('_preprocessing_for_prophet')
+        tsa_logger.debug(df)
         return df
 
     def _make_forecast(self) -> pd.DataFrame:
@@ -108,12 +112,16 @@ class MyProphet:
 
         # 향후 180일 동안의 주가 예측
         future = self.model.make_future_dataframe(periods=180)
+        tsa_logger.debug('_make_forecast_future')
+        tsa_logger.debug(future)
 
         # 미래 데이터에 거래량 추가 (평균 거래량을 사용해 정규화)
         future_volume = pd.DataFrame({'volume': [self.raw_data['Volume'].mean()] * len(future)})
         future['volume_scaled'] = self.scaler.transform(future_volume[['volume']])
 
         forecast = self.model.predict(future)
+        tsa_logger.debug('_make_forecast')
+        tsa_logger.debug(forecast)
         return forecast
 
     def get_yhat(self) -> dict:
@@ -290,13 +298,16 @@ class MyLSTM:
 
         # 4년 전 날짜 계산 (4년 = 365일 * 4)
         four_years_ago = today - timedelta(days=365 * 4)
-        tsa_logger.info(f'start: {four_years_ago.strftime('%Y-%m-%d')}, end: {today.strftime('%Y-%m-%d')}')
+        tsa_logger.info(f"start: {four_years_ago.strftime('%Y-%m-%d')}, end: {today.strftime('%Y-%m-%d')}")
 
-        return yf.download(
+        df = yf.download(
             self.code + '.KS',
             start=four_years_ago.strftime('%Y-%m-%d'),
             end=today.strftime('%Y-%m-%d')
         )
+        df.index = df.index.tz_localize(None)
+        tsa_logger.debug(df)
+        return df
 
     def _preprocessing_for_lstm(self) -> LSTMData:
         """
@@ -538,17 +549,61 @@ class MyLSTM:
         future_dates, final_future_predictions = self.get_final_predictions(refresh=refresh)
         final_future_predictions = final_future_predictions.reshape(-1) # 차원을 하나 줄인다.
 
-        # Plotly를 사용한 시각화
+        # 데이터 준비
+        self.raw_data = self.raw_data.reset_index()
+        data = self.raw_data[['Date', 'Close']][-120:].reset_index(drop=True)
+
+        # 'Date'와 'Close' 열 추출
+        actual_dates = pd.to_datetime(data['Date'])
+        actual_close = data['Close']
+
+        # 'actual_close'가 Series인지 확인
+        if isinstance(actual_close, pd.DataFrame):
+            actual_close = actual_close.squeeze()
+
+        # 'Close' 열의 데이터 타입 확인
+        actual_close = actual_close.astype(float)
+
+        # 예측 데이터 준비
+        predicted_dates = pd.to_datetime(future_dates)
+        predicted_close = pd.Series(final_future_predictions, index=range(len(final_future_predictions))).astype(float)
+
+        # 그래프 생성
         fig = go.Figure()
 
-        # 실제 데이터
-        fig.add_trace(go.Scatter(x=self.raw_data.index[-120:], y=self.raw_data['Close'][-120:], mode='markers', name='실제주가'))
-        tsa_logger.debug(f"self.raw_data.index[-120:] - {self.raw_data.index[-120:]}")
-        tsa_logger.debug(f"self.raw_data['Close'][-120:] - {self.raw_data['Close'][-120:]}")
-        # 예측 데이터
-        fig.add_trace(go.Scatter(x=future_dates, y=final_future_predictions, mode='lines+markers', name='예측치(30일)'))
-        tsa_logger.debug(f"future_dates - {future_dates}")
-        tsa_logger.debug(f"final_future_predictions - {final_future_predictions}")
+        # 실제 데이터 추가
+        fig.add_trace(go.Scatter(
+            x=actual_dates,
+            y=actual_close,
+            mode='markers',
+            name='실제주가'
+        ))
+
+        # 예측 데이터 추가
+        fig.add_trace(go.Scatter(
+            x=predicted_dates,
+            y=predicted_close,
+            mode='lines+markers',
+            name='예측치(30일)'
+        ))
+
+        # 레이아웃 업데이트
+        fig.update_layout(
+            xaxis_title='일자',
+            yaxis_title='주가(원)',
+            xaxis=dict(
+                tickformat='%Y/%m',
+            ),
+            yaxis=dict(
+                tickformat=".0f",
+            ),
+            showlegend=True,
+        )
+
+        tsa_logger.debug(f"actual_dates({len(actual_dates)}) - {actual_dates}")
+        tsa_logger.debug(f"actual_close({len(actual_close)} - {actual_close}")
+        tsa_logger.debug(f"predicted_dates({len(future_dates)}) - {future_dates}")
+        tsa_logger.debug(f"predicted_close({len(predicted_close)}) - {predicted_close}")
 
         fig.update_layout(
             # title=f'{self.code} {self.name} 주가 예측 그래프(prophet)',

{analyser_hj3415-2.9.3 → analyser_hj3415-2.9.5}/pyproject.toml

@@ -5,7 +5,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "analyser_hj3415"
-version = "2.9.3"
+version = "2.9.5"
 description = "Stock analyser and database processing programs"
 readme = "README.md"
 requires-python = ">=3.6"
@@ -28,6 +28,7 @@ dependencies = [
     "prophet>=1.1.6",
     "kaleido>=0.2.1",   #plotly로 이미지출력위해
     "matplotlib>=3.9.2",
+    "tensorflow>=2.18.0",
 ]
 
 [project.scripts]