analyser_hj3415 2.7.8__tar.gz → 2.8.1__tar.gz

{analyser_hj3415-2.7.8 → analyser_hj3415-2.8.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: analyser_hj3415
-Version: 2.7.8
+Version: 2.8.1
 Summary: Stock analyser and database processing programs
 Author-email: Hyungjin Kim <hj3415@gmail.com>
 Description-Content-Type: text/markdown

analyser_hj3415-2.7.8/analyser_hj3415/myprophet.py → analyser_hj3415-2.8.1/analyser_hj3415/tsa.py

@@ -1,24 +1,36 @@
+"""
+Time Series Analysis
+"""
 import yfinance as yf
 from datetime import datetime, timedelta
 import pandas as pd
 from prophet import Prophet
 from sklearn.preprocessing import StandardScaler
-from utils_hj3415 import utils
+from utils_hj3415 import utils, helpers
 from typing import Optional
 import plotly.graph_objs as go
 from plotly.offline import plot
 import matplotlib.pyplot as plt  # Matplotlib 수동 임포트
 from db_hj3415 import myredis
+from collections import OrderedDict
+from analyser_hj3415 import eval
+
+import logging
+analyser_logger = helpers.setup_logger('analyser_logger', logging.WARNING)
+
+expire_time = 3600 * 12
 
 class MyProphet:
     def __init__(self, code: str):
         assert utils.is_6digit(code), f'Invalid value : {code}'
-        self._code = code
-        self.name = myredis.Corps(code, 'c101').get_name()
-        self.raw_data = self.get_raw_data()
-
         self.scaler = StandardScaler()
+
         self.model = Prophet()
+        self._code = code
+        self.name = myredis.Corps(code, 'c101').get_name()
+        self.raw_data = self._get_raw_data()
+        self.df_real = self._preprocessing_for_prophet()
+        self.df_forecast = self._make_forecast()
 
     @property
     def code(self) -> str:
@@ -27,11 +39,25 @@ class MyProphet:
     @code.setter
     def code(self, code: str):
         assert utils.is_6digit(code), f'Invalid value : {code}'
+        analyser_logger.info(f'change code : {self.code} -> {code}')
+        self.model = Prophet()
         self._code = code
         self.name = myredis.Corps(code, 'c101').get_name()
-        self.raw_data = self.get_raw_data()
-
-    def get_raw_data(self) -> pd.DataFrame:
+        self.raw_data = self._get_raw_data()
+        self.df_real = self._preprocessing_for_prophet()
+        self.df_forecast = self._make_forecast()
+
+    @staticmethod
+    def is_valid_date(date_string):
+        try:
+            # %Y-%m-%d 형식으로 문자열을 datetime 객체로 변환 시도
+            datetime.strptime(date_string, '%Y-%m-%d')
+            return True
+        except ValueError:
+            # 변환이 실패하면 ValueError가 발생, 형식이 맞지 않음
+            return False
+
+    def _get_raw_data(self) -> pd.DataFrame:
         """
         야후에서 해당 종목의 4년간 주가 raw data를 받아온다.
         :return:
@@ -48,7 +74,7 @@ class MyProphet:
             end=today.strftime('%Y-%m-%d')
         )
 
-    def preprocessing_for_prophet(self) -> pd.DataFrame:
+    def _preprocessing_for_prophet(self) -> pd.DataFrame:
         """
         Prophet이 사용할 수 있도록 데이터 준비
         ds는 날짜, y는 주가
@@ -61,14 +87,11 @@ class MyProphet:
         df['volume_scaled'] = self.scaler.fit_transform(df[['volume']])
         return df
 
-    def make_forecast(self) -> pd.DataFrame:
-        # Prophet을 위한 dataframe 만들기
-        df_for_prophet = self.preprocessing_for_prophet()
-
+    def _make_forecast(self) -> pd.DataFrame:
         # 정규화된 'volume_scaled' 변수를 외부 변수로 추가
         self.model.add_regressor('volume_scaled')
 
-        self.model.fit(df_for_prophet)
+        self.model.fit(self.df_real)
 
         # 향후 180일 동안의 주가 예측
         future = self.model.make_future_dataframe(periods=180)
@@ -80,30 +103,37 @@ class MyProphet:
         forecast = self.model.predict(future)
         return forecast
 
+    def get_yhat(self) -> dict:
+        """
+        최근 날짜의 예측데이터를 반환한다.
+        :return: {'ds':..., 'yhat':.., 'yhat_lower':.., 'yhat_upper':..,}
+        """
+        df = self.df_forecast
+        last_real_date = self.df_real.iloc[-1]['ds']
+        analyser_logger.info(last_real_date)
+        yhat_dict = df[df['ds']==last_real_date].iloc[0][['ds', 'yhat_lower', 'yhat_upper', 'yhat']].to_dict()
+        analyser_logger.info(yhat_dict)
+        return yhat_dict
+
     def export_to(self, to="str") -> Optional[str]:
         """
         prophet과 plotly로 그래프를 그려서 html을 문자열로 반환
         :param to: str, png, htmlfile, show
         :return:
         """
-        # 실제 데이터
-        df = self.preprocessing_for_prophet()
-        # 예측 데이터
-        forecast = self.make_forecast()
-
         # Plotly를 사용한 시각화
         fig = go.Figure()
 
         # 실제 데이터
-        fig.add_trace(go.Scatter(x=df['ds'], y=df['y'], mode='markers', name='실제주가'))
+        fig.add_trace(go.Scatter(x=self.df_real['ds'], y=self.df_real['y'], mode='markers', name='실제주가'))
         # 예측 데이터
-        fig.add_trace(go.Scatter(x=forecast['ds'], y=forecast['yhat'], mode='lines', name='예측치'))
+        fig.add_trace(go.Scatter(x=self.df_forecast['ds'], y=self.df_forecast['yhat'], mode='lines', name='예측치'))
 
         # 상한/하한 구간
         fig.add_trace(
-            go.Scatter(x=forecast['ds'], y=forecast['yhat_upper'], fill=None, mode='lines', name='상한'))
+            go.Scatter(x=self.df_forecast['ds'], y=self.df_forecast['yhat_upper'], fill=None, mode='lines', name='상한'))
         fig.add_trace(
-            go.Scatter(x=forecast['ds'], y=forecast['yhat_lower'], fill='tonexty', mode='lines', name='하한'))
+            go.Scatter(x=self.df_forecast['ds'], y=self.df_forecast['yhat_lower'], fill='tonexty', mode='lines', name='하한'))
 
         fig.update_layout(
             title=f'{self.code} {self.name} 주가 예측 그래프(prophet)',
@@ -130,15 +160,48 @@ class MyProphet:
             return None
         elif to == 'show':
             # 예측 결과 출력
-            print(forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail())
+            print(self.df_forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail())
             # 예측 결과 시각화 (Matplotlib 사용)
-            fig = self.model.plot(forecast)
+            fig = self.model.plot(self.df_forecast)
             # 추세 및 계절성 시각화
-            fig2 = self.model.plot_components(forecast)
+            fig2 = self.model.plot_components(self.df_forecast)
             plt.show()  # 시각화 창 띄우기
         else:
             Exception("to 인자가 맞지 않습니다.")
 
+    @classmethod
+    def ranking(cls, refresh = False) -> OrderedDict:
+        """
+        가장 최근 날짜의 랭킹 분석
+        :param refresh:
+        :return:
+        """
+        print("**** Start myprophet_ranking... ****")
+        redis_name = 'myprophet_ranking'
+
+        print(
+            f"redisname: '{redis_name}' / refresh : {refresh} / expire_time : {expire_time / 3600}h")
+
+        def fetch_ranking() -> dict:
+            data = {}
+            p = MyProphet('005930')
+            for i, code in enumerate(myredis.Corps.list_all_codes()):
+                p.code = code
+                last_real_data = p._preprocessing_for_prophet().iloc[-1]
+                recent_price = last_real_data['y']
+                recent_date = datetime.strftime(last_real_data['ds'], '%Y-%m-%d')
+                yhat_dict = p.get_yhat()
+                analyser_logger.info(f'recent_price: {recent_price}, yhat_dict: {yhat_dict}')
+                yhat_lower = int(yhat_dict['yhat_lower'])
+                if recent_price < yhat_lower:
+                    deviation = int(eval.Tools.cal_deviation(recent_price, yhat_lower))
+                    data[code] = deviation
+                    print(f"{i}.{p.code}/{p.name} date: {recent_date} 가격:{recent_price} 기대하한값:{yhat_lower} 편차:{deviation}")
+            return data
+
+        data_dict = myredis.Base.fetch_and_cache_data(redis_name, refresh, fetch_ranking, timer=expire_time)
+
+        return OrderedDict(sorted(data_dict.items(), key=lambda item: item[1], reverse=True))
 
 
 

{analyser_hj3415-2.7.8 → analyser_hj3415-2.8.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "analyser_hj3415"
-version = "2.7.8"
+version = "2.8.1"
 authors = [{name = "Hyungjin Kim", email = "hj3415@gmail.com"}]
 description = "Stock analyser and database processing programs"
 readme = "README.md"