Update app.py

app.py

@@ -1,20 +1,21 @@
 import streamlit as st
 from datetime import date
 import yfinance as yf
-import pandas as pd  # Importing pandas to avoid NameError
+import pandas as pd
 from prophet import Prophet
 from prophet.plot import plot_plotly
-from plotly import graph_objs as go
+import plotly.graph_objects as go
+from sklearn.metrics import mean_absolute_error, mean_squared_error
 
 # Constants for date range
 START = "2015-01-01"
 TODAY = date.today().strftime("%Y-%m-%d")
 
 # Streamlit app title
-st.title('Stock & Cryptocurrency Forecast App')
+st.title('Advanced Stock & Cryptocurrency Forecast App')
 
 # Stock and cryptocurrency selection
-assets = ('GOOG', 'AAPL', 'MSFT', 'GME', 'BTC-USD', 'ETH-USD')  # Added BTC and ETH
+assets = ('GOOG', 'AAPL', 'MSFT', 'GME', 'BTC-USD', 'ETH-USD')
 selected_asset = st.selectbox('Select dataset for prediction', assets)
 
 # Years of prediction slider
@@ -39,38 +40,24 @@ st.write(data.tail())
 
 # Check if 'Close' column exists before converting
 if 'Close' in data.columns:
-    # Check data types of the DataFrame
-    st.write("Data Types:")
-    st.write(data.dtypes)
-
-    # Display unique values in the 'Close' column for debugging (accessing directly)
-    unique_close_values = data['Close'].unique()  # Accessing directly without pd.Series
-    st.write("Unique values in 'Close' column:")
-    st.write(unique_close_values)
-
     # Ensure 'Close' prices are numeric and handle any missing values
     try:
-        # Convert to numeric with coercion for invalid parsing
         data['Close'] = pd.to_numeric(data['Close'], errors='coerce')
-        # Drop rows with NaN values in 'Close'
         data.dropna(subset=['Close'], inplace=True)
     except Exception as e:
         st.error(f"Error converting 'Close' prices to numeric: {e}")
-else:
-    st.error("The 'Close' column is missing from the data. Please check the selected asset.")
 
-# Check if there is valid data after processing
-if not data.empty and 'Close' in data.columns:
-    # Plot raw data function
-    def plot_raw_data():
-        fig = go.Figure()
-        fig.add_trace(go.Scatter(x=data['Date'], y=data['Open'], name="Open Price", line=dict(color='blue')))
-        fig.add_trace(go.Scatter(x=data['Date'], y=data['Close'], name="Close Price", line=dict(color='orange')))
-        fig.layout.update(title_text='Time Series Data with Rangeslider', xaxis_rangeslider_visible=True)
-        st.plotly_chart(fig)
-
-    # Call the plotting function if there's enough data
-    plot_raw_data()
+    # Plot raw data using candlestick chart for better visualization
+    fig_candlestick = go.Figure(data=[go.Candlestick(x=data['Date'],
+                                                       open=data['Open'],
+                                                       high=data['High'],
+                                                       low=data['Low'],
+                                                       close=data['Close'])])
+    fig_candlestick.update_layout(title=f'{selected_asset} Price Data',
+                                   xaxis_title='Date',
+                                   yaxis_title='Price (USD)',
+                                   xaxis_rangeslider_visible=True)
+    st.plotly_chart(fig_candlestick)
 
     # Prepare data for Prophet model
     df_train = data[['Date', 'Close']]
@@ -78,26 +65,34 @@ if not data.empty and 'Close' in data.columns:
 
     # Create and fit the Prophet model
     m = Prophet()
-    
+
     try:
         m.fit(df_train)
-        
+
         # Create future dataframe and make predictions
         future = m.make_future_dataframe(periods=period)
         forecast = m.predict(future)
 
+        # Calculate performance metrics on historical predictions (if available)
+        if len(df_train) > 30:  # Ensure we have enough historical data to evaluate
+            historical_forecast = m.predict(df_train)
+            mae = mean_absolute_error(df_train['y'], historical_forecast['yhat'])
+            rmse = mean_squared_error(df_train['y'], historical_forecast['yhat', squared=False])
+            st.write(f"Mean Absolute Error (MAE): {mae:.2f}")
+            st.write(f"Root Mean Squared Error (RMSE): {rmse:.2f}")
+
         # Show forecast data and plot forecast
         st.subheader('Forecast Data')
         st.write(forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail())
-        st.write(f'Forecast plot for the next {n_years} years')
-
-        fig1 = plot_plotly(m, forecast)
-        st.plotly_chart(fig1)
+        
+        # Plot forecast with confidence intervals
+        fig_forecast = plot_plotly(m, forecast)
+        st.plotly_chart(fig_forecast)
 
         # Show forecast components
         st.subheader("Forecast Components")
-        fig2 = m.plot_components(forecast)
-        st.plotly_chart(fig2)
+        fig_components = m.plot_components(forecast)
+        st.plotly_chart(fig_components)
 
         # Additional Insights: Displaying key metrics
         st.subheader("Key Metrics")