graph on bitcoin

app.py

@@ -128,7 +128,8 @@ def index():
                 df_prophet_with_date['ds_date'] = df_prophet_with_date['ds'].dt.normalize()
                 temp_merged = pd.merge(df_prophet_with_date, df_fng, on='ds_date', how='left')
                 temp_merged['fear_greed'] = temp_merged['fear_greed'].ffill().bfill()
-                temp_merged['fear_greed'].fillna(50, inplace=True) # Final fallback
+                # FIX for Pandas FutureWarning: Use direct assignment instead of inplace=True on potentially chained access
+                temp_merged['fear_greed'] = temp_merged['fear_greed'].fillna(50) # Final fallback
                 df_merged = temp_merged[['ds', 'y', 'volume', 'fear_greed']].copy()
                 fng_available = True
                 data_confidence_score += 25
@@ -156,21 +157,25 @@ def index():
         try:
             print("--- Instantiating Prophet model ---")
             model = Prophet(interval_width=0.95) # 95% prediction interval
+            volume_regressor_added = False # <<< FIX: Initialize flag
+            fng_regressor_added = False    # <<< FIX: Initialize flag
             print("--- Adding Regressors ---")
 
             # Add 'volume' regressor if valid
             if 'volume' in df_merged.columns and df_merged['volume'].notna().any():
                  model.add_regressor('volume')
+                 volume_regressor_added = True # <<< FIX: Set flag
                  print("Added 'volume' regressor.")
             else: print("--- Skipping 'volume' regressor (column missing or all NaN) ---")
 
             # Add 'fear_greed' regressor if valid and available
             if fng_available and 'fear_greed' in df_merged.columns and df_merged['fear_greed'].notna().any():
                  model.add_regressor('fear_greed')
+                 fng_regressor_added = True # <<< FIX: Set flag
                  print("Added 'fear_greed' regressor.")
             else:
+                 fng_available = False # Ensure state consistency
                  print("Fitting model WITHOUT 'fear_greed' regressor (not available or all NaN).")
-                 fng_available = False
 
             # Final check on data points before fitting
             df_merged.dropna(subset=['y'], inplace=True) # Drop rows if 'y' became NaN
@@ -184,12 +189,14 @@ def index():
             future = model.make_future_dataframe(periods=24, freq='h')
 
             print("--- Populating future regressor values... ---")
-            if 'volume' in model.regressors:
+            # <<< FIX: Use flags instead of checking model.regressors >>>
+            if volume_regressor_added:
                 last_volume = df_merged['volume'].iloc[-1] if not df_merged['volume'].empty else 0
                 future['volume'] = last_volume
                 print(f"   - Set future 'volume' to: {last_volume}")
 
-            if 'fear_greed' in model.regressors:
+            # <<< FIX: Use flags instead of checking model.regressors >>>
+            if fng_regressor_added:
                 if current_fng_value is not None and pd.notna(current_fng_value):
                     future['fear_greed'] = float(current_fng_value)
                     used_current_fng_for_future = True
@@ -376,4 +383,3 @@ if __name__ == '__main__':
     # Port 5000 is default, change if needed
     # Debug=True provides more error detail and auto-reloads on code changes
     app.run(debug=True, host='0.0.0.0', port=5000)
\ No newline at end of file
-

templates/dashboard.html

@@ -181,7 +181,7 @@
             }
 
             // Chart Rendering
-            const plotData = {{ plot_data | tojson | safe }}; // Embed data from Flask
+            const plotData = JSON.parse('{{ plot_data | tojson | safe }}'); // Embed data from Flask
             const chartCanvas = document.getElementById('priceChart');
 
             if (plotData && plotData.labels && plotData.labels.length > 0 && chartCanvas) {