Update app.py

app.py

@@ -1,10 +1,14 @@
 import gradio as gr
 import yfinance as yf
 from prophet import Prophet
+from sklearn.linear_model import LinearRegression
 import pandas as pd
 from datetime import datetime
 import plotly.graph_objects as go
 
+def predict_future_prices(ticker, periods=1825):
+data = download_data(ticker)
+
 def download_data(ticker, start_date='2010-01-01'):
     """ 주식 데이터를 다운로드하고 포맷을 조정하는 함수 """
     data = yf.download(ticker, start=start_date)
@@ -20,37 +24,46 @@ def download_data(ticker, start_date='2010-01-01'):
 
 def predict_future_prices(ticker, periods=1825):
     data = download_data(ticker)
-    
     # Prophet 모델 생성 및 학습
-    model = Prophet(daily_seasonality=False, weekly_seasonality=False, yearly_seasonality=True)
-    model.fit(data)
-    
+    model_prophet = Prophet(daily_seasonality=False, weekly_seasonality=False, yearly_seasonality=True)
+    model_prophet.fit(data)
+    # 미래 데이터 프레임 생성 및 예측
+    future = model_prophet.make_future_dataframe(periods=periods, freq='D')
+    forecast_prophet = model_prophet.predict(future)
+    # Linear Regression 모델 생성 및 학습
+    model_lr = LinearRegression()
+    X = pd.to_numeric(pd.Series(range(len(data))))
+    y = data['y'].values
+    model_lr.fit(X.values.reshape(-1, 1), y)
     # 미래 데이터 프레임 생성 및 예측
-    future = model.make_future_dataframe(periods=periods, freq='D')
-    forecast = model.predict(future)
-    
+    future_dates = pd.date_range(start=data['ds'].iloc[-1], periods=periods+1, freq='D')[1:]
+    future_lr = pd.DataFrame({'ds': future_dates})
+    future_lr['ds'] = future_lr['ds'].dt.strftime('%Y-%m-%d')
+    X_future = pd.to_numeric(pd.Series(range(len(data), len(data) + len(future_lr))))
+    future_lr['yhat'] = model_lr.predict(X_future.values.reshape(-1, 1))
     # 예측 결과 그래프 생성
-    forecast['ds'] = forecast['ds'].dt.strftime('%Y-%m-%d')
+    forecast_prophet['ds'] = forecast_prophet['ds'].dt.strftime('%Y-%m-%d')
     fig = go.Figure()
-    fig.add_trace(go.Scatter(x=forecast['ds'], y=forecast['yhat'], mode='lines', name='Forecast (Blue)'))
+    fig.add_trace(go.Scatter(x=forecast_prophet['ds'], y=forecast_prophet['yhat'], mode='lines', name='Prophet Forecast (Blue)'))
+    fig.add_trace(go.Scatter(x=future_lr['ds'], y=future_lr['yhat'], mode='lines', name='Linear Regression Forecast (Red)', line=dict(color='red')))
     fig.add_trace(go.Scatter(x=data['ds'], y=data['y'], mode='lines', name='Actual (Black)', line=dict(color='black')))
-    
-    return fig, forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']]
+    return fig, forecast_prophet[['ds', 'yhat', 'yhat_lower', 'yhat_upper']], future_lr[['ds', 'yhat']]
 
-# Gradio 인터페이스 설정 및 실행
-with gr.Blocks() as app:
+css = """footer { visibility: hidden; }"""
+
+with gr.Blocks(css=css) as app:
     with gr.Row():
         ticker_input = gr.Textbox(value="AAPL", label="Enter Stock Ticker for Forecast")
         periods_input = gr.Number(value=1825, label="Forecast Period (days)")
         forecast_button = gr.Button("Generate Forecast")
-        
-    forecast_chart = gr.Plot(label="Forecast Chart")
-    forecast_data = gr.Dataframe(label="Forecast Data")
-    
+        forecast_chart = gr.Plot(label="Forecast Chart")
+        forecast_data_prophet = gr.Dataframe(label="Prophet Forecast Data")
+        forecast_data_lr = gr.Dataframe(label="Linear Regression Forecast Data")
+
     forecast_button.click(
         fn=predict_future_prices,
         inputs=[ticker_input, periods_input],
-        outputs=[forecast_chart, forecast_data]
+        outputs=[forecast_chart, forecast_data_prophet, forecast_data_lr]
     )
 
 app.launch()