Update app.py

app.py

@@ -56,36 +56,64 @@ class SentimentAnalyzerApp:
         
     def load_sample_data(self):
         """Create sample data for demo purposes"""
-        sample_data = {
-            'date': ['2024-01-01', '2024-01-02', '2024-01-03', '2024-01-04', '2024-01-05'],
-            'review': [
-                'This app is absolutely amazing and very helpful!',
-                'The application works okay but could be better.',
-                'I am very disappointed with the performance.',
-                'Excellent features and great user interface.',
-                'Not what I expected, needs improvement.'
-            ],
-            'rating': [5, 3, 1, 5, 2],
-            'platform': ['Web', 'Mobile', 'Web', 'Mobile', 'Web'],
-            'language': ['en', 'en', 'en', 'en', 'en'],
-            'location': ['USA', 'UK', 'Canada', 'Australia', 'India'],
-            'verified_purchase': ['Yes', 'No', 'Yes', 'Yes', 'No'],
-            'helpful_votes': [10, 2, 5, 8, 1]
-        }
-        self.df = pd.DataFrame(sample_data)
-        self.df['date'] = pd.to_datetime(self.df['date'])
-        
-        # Create sentiment labels
-        def get_sentiment(rating):
-            if rating >= 4:
-                return 'Positive'
-            elif rating == 3:
-                return 'Neutral'
-            else:
-                return 'Negative'
-        
-        self.df['sentiment'] = self.df['rating'].apply(get_sentiment)
-        return True
+        try:
+            sample_data = {
+                'date': ['2024-01-01', '2024-01-02', '2024-01-03', '2024-01-04', '2024-01-05'],
+                'review': [
+                    'This app is absolutely amazing and very helpful!',
+                    'The application works okay but could be better.',
+                    'I am very disappointed with the performance.',
+                    'Excellent features and great user interface.',
+                    'Not what I expected, needs improvement.'
+                ],
+                'rating': [5, 3, 1, 5, 2],
+                'platform': ['Web', 'Mobile', 'Web', 'Mobile', 'Web'],
+                'language': ['en', 'en', 'en', 'en', 'en'],
+                'location': ['USA', 'UK', 'Canada', 'Australia', 'India'],
+                'verified_purchase': ['Yes', 'No', 'Yes', 'Yes', 'No'],
+                'helpful_votes': [10, 2, 5, 8, 1]
+            }
+            self.df = pd.DataFrame(sample_data)
+            self.df['date'] = pd.to_datetime(self.df['date'])
+            
+            # Create sentiment labels
+            def get_sentiment(rating):
+                if rating >= 4:
+                    return 'Positive'
+                elif rating == 3:
+                    return 'Neutral'
+                else:
+                    return 'Negative'
+            
+            self.df['sentiment'] = self.df['rating'].apply(get_sentiment)
+            return True
+        except Exception as e:
+            st.error(f"Error creating sample data: {e}")
+            return False
+    
+    def load_real_data(self):
+        """Try to load real data from file"""
+        try:
+            data_path = 'data/chatgpt_style_reviews_dataset.csv'
+            if os.path.exists(data_path):
+                self.df = pd.read_csv(data_path)
+                self.df['date'] = pd.to_datetime(self.df['date'], errors='coerce')
+                
+                # Create sentiment labels
+                def get_sentiment(rating):
+                    if rating >= 4:
+                        return 'Positive'
+                    elif rating == 3:
+                        return 'Neutral'
+                    else:
+                        return 'Negative'
+                
+                self.df['sentiment'] = self.df['rating'].apply(get_sentiment)
+                return True
+            return False
+        except Exception as e:
+            st.error(f"Error loading real data: {e}")
+            return False
     
     def load_model(self):
         """Try to load model, but use simulated predictions if not available"""
@@ -104,6 +132,14 @@ class SentimentAnalyzerApp:
             st.warning(f"Model loading failed: {e}. Using simulated mode.")
             return False
     
+    def ensure_data_loaded(self):
+        """Ensure data is loaded, use sample if real data not available"""
+        if self.df is None:
+            # First try to load real data
+            if not self.load_real_data():
+                # If real data fails, load sample data
+                self.load_sample_data()
+    
     def predict_sentiment(self, text):
         """Predict sentiment for new text"""
         if self.model is not None and self.vectorizer is not None:
@@ -174,9 +210,7 @@ class SentimentAnalyzerApp:
         st.markdown("### Customer Review Sentiment Analysis Dashboard")
         
         # Initialize and load data
-        if 'data_loaded' not in st.session_state:
-            self.load_sample_data()
-            st.session_state.data_loaded = True
+        self.ensure_data_loaded()
         
         if 'model_loaded' not in st.session_state:
             st.session_state.model_loaded = self.load_model()
@@ -202,6 +236,9 @@ class SentimentAnalyzerApp:
         """Overview page"""
         st.header("📊 Project Overview")
         
+        # Ensure data is loaded
+        self.ensure_data_loaded()
+        
         # Key metrics
         col1, col2, col3, col4 = st.columns(4)
         
@@ -241,7 +278,11 @@ class SentimentAnalyzerApp:
                         title='Sentiment Distribution')
             st.plotly_chart(fig, use_container_width=True)
         
-        st.info("💡 This is a demo with sample data. Upload your dataset to the 'data' folder for real analysis.")
+        # Show data source info
+        if hasattr(self, 'using_real_data') and self.using_real_data:
+            st.success("✅ Using real dataset from file")
+        else:
+            st.info("💡 Using sample data for demo. Upload your dataset to the 'data' folder for real analysis.")
     
     def show_model_demo(self):
         """Interactive model demo"""
@@ -307,13 +348,19 @@ class SentimentAnalyzerApp:
         for i, example in enumerate(examples):
             with cols[i % 3]:
                 if st.button(f"'{example[:30]}...'", use_container_width=True):
-                    st.session_state.demo_text = example
                     st.rerun()
 
     def show_analysis(self):
         """Analysis page"""
         st.header("📈 Data Analysis")
         
+        # Ensure data is loaded
+        self.ensure_data_loaded()
+        
+        if self.df is None:
+            st.error("No data available for analysis.")
+            return
+        
         # Platform analysis
         st.subheader("Platform Comparison")
         platform_counts = self.df['platform'].value_counts()
@@ -362,6 +409,13 @@ class SentimentAnalyzerApp:
         """Insights page"""
         st.header("💡 Business Insights & Recommendations")
         
+        # Ensure data is loaded
+        self.ensure_data_loaded()
+        
+        if self.df is None:
+            st.error("No data available for insights.")
+            return
+        
         # Key metrics
         positive_pct = (self.df['sentiment'] == 'Positive').mean() * 100
         avg_rating = self.df['rating'].mean()