Update app.py

app.py

@@ -1,3 +1,14 @@
+import sys
+import subprocess
+
+# Check and install required libraries
+required_libraries = ["streamlit", "transformers", "yfinance", "matplotlib", "Pillow"]
+for lib in required_libraries:
+    try:
+        __import__(lib)
+    except ImportError:
+        subprocess.check_call([sys.executable, "-m", "pip", "install", lib])
+
 import streamlit as st
 from transformers import pipeline
 from PIL import Image
@@ -5,11 +16,17 @@ import yfinance as yf
 import matplotlib.pyplot as plt
 import io
 
-# Install yfinance if not already installed
-subprocess.check_call([sys.executable, "-m", "pip", "install", "yfinance"])
+# Cache the stock data retrieval
+@st.cache_data
+def get_stock_data(ticker, start_date, end_date):
+    return yf.download(ticker, start=start_date, end=end_date)
 
 # Load the image classification pipeline
-classifier = pipeline(task="image-classification", model="julien-c/bullish-or-bearish")
+try:
+    classifier = pipeline(task="image-classification", model="microsoft/resnet-50")
+except Exception as e:
+    st.error(f"Error loading the model: {e}")
+    st.stop()
 
 # Set up the Streamlit app title
 st.title("Stock Trend Analyzer: Bullish or Bearish? 📈📉")
@@ -24,22 +41,23 @@ end_date = col2.date_input("End date")
 
 if st.button("Analyze"):
     # Fetch stock data
-    stock_data = yf.download(ticker, start=start_date, end=end_date)
+    with st.spinner("Fetching stock data..."):
+        stock_data = get_stock_data(ticker, start_date, end_date)
     
     if stock_data.empty:
         st.error("No data available for the selected stock and date range.")
     else:
         # Create a plot of the stock data
-        plt.figure(figsize=(10, 6))
-        plt.plot(stock_data.index, stock_data['Close'])
-        plt.title(f"{ticker} Stock Price")
-        plt.xlabel("Date")
-        plt.ylabel("Price")
-        plt.grid(True)
+        fig, ax = plt.subplots(figsize=(10, 6))
+        ax.plot(stock_data.index, stock_data['Close'])
+        ax.set_title(f"{ticker} Stock Price")
+        ax.set_xlabel("Date")
+        ax.set_ylabel("Price")
+        ax.grid(True)
         
         # Save the plot to a buffer
         buf = io.BytesIO()
-        plt.savefig(buf, format='png')
+        fig.savefig(buf, format='png')
         buf.seek(0)
         
         # Convert the buffer to an image
@@ -54,8 +72,8 @@ if st.button("Analyze"):
         
         # Display the predictions
         st.header("Trend Analysis")
-        for p in predictions:
-            sentiment = p['label'].split('_')[0].capitalize()
+        for p in predictions[:2]:  # Display top 2 predictions
+            sentiment = "Bullish" if "up" in p['label'].lower() else "Bearish"
             confidence = round(p['score'] * 100, 1)
             st.subheader(f"{sentiment}: {confidence}%")
             
@@ -79,4 +97,7 @@ if st.button("Analyze"):
             st.info("The stock price remained unchanged over the selected period.")
 
 else:
-    st.info("Enter a stock ticker and select a date range to analyze the trend.")
+    st.info("Enter a stock ticker and select a date range to analyze the trend.")
+
+# Clean up matplotlib figures
+plt.close('all')