feat: filter out non-positive APR values and update graph handling for positive values only

app.py

@@ -251,15 +251,15 @@ def fetch_apr_data_from_db():
                 # Add is_dummy flag (all real data)
                 apr_data["is_dummy"] = False
                 
-                # Mark negative values as "Performance" metrics
-                if apr_data["apr"] < 0:
-                    apr_data["metric_type"] = "Performance"
-                    logger.debug(f"Agent {agent_name} ({attr['agent_id']}): Performance value: {apr_data['apr']}")
-                else:
+                # CHANGE: Only include positive APR values (greater than 0)
+                if apr_data["apr"] > 0:
                     apr_data["metric_type"] = "APR"
                     logger.debug(f"Agent {agent_name} ({attr['agent_id']}): APR value: {apr_data['apr']}")
-                    
-                apr_data_list.append(apr_data)
+                    # Add to the data list only if value is positive
+                    apr_data_list.append(apr_data)
+                else:
+                    # Log that we're skipping non-positive values
+                    logger.debug(f"Skipping non-positive value for agent {agent_name} ({attr['agent_id']}): {apr_data['apr']}")
         
         # Convert list of dictionaries to DataFrame
         if not apr_data_list:
@@ -273,7 +273,8 @@ def fetch_apr_data_from_db():
         logger.info(f"Created DataFrame with {len(global_df)} rows")
         logger.info(f"DataFrame columns: {global_df.columns.tolist()}")
         logger.info(f"APR statistics: min={global_df['apr'].min()}, max={global_df['apr'].max()}, mean={global_df['apr'].mean()}")
-        logger.info(f"Metric types count: {global_df['metric_type'].value_counts().to_dict()}")
+        # After filtering, all values are APR type
+        logger.info("All values are APR type (positive values only)")
         logger.info(f"Agents count: {global_df['agent_name'].value_counts().to_dict()}")
         
         # Log the entire dataframe for debugging
@@ -488,7 +489,7 @@ def create_combined_time_series_graph(df):
     # CRITICAL: Log the exact dataframe we're using for plotting to help debug
     logger.info(f"Graph data - shape: {df.shape}, columns: {df.columns}")
     logger.info(f"Graph data - unique agents: {df['agent_name'].unique().tolist()}")
-    logger.info(f"Graph data - unique metric types: {df['metric_type'].unique().tolist()}")
+    logger.info("Graph data - all positive APR values only")
     logger.info(f"Graph data - min APR: {df['apr'].min()}, max APR: {df['apr'].max()}")
     
     # Export full dataframe to CSV for debugging
@@ -514,19 +515,13 @@ def create_combined_time_series_graph(df):
             f.write(f"  Total data points: {len(agent_data)}\n")
             
             apr_data = agent_data[agent_data['metric_type'] == 'APR']
-            perf_data = agent_data[agent_data['metric_type'] == 'Performance']
             
             f.write(f"  APR data points: {len(apr_data)}\n")
-            f.write(f"  Performance data points: {len(perf_data)}\n")
             
             if not apr_data.empty:
                 f.write(f"  APR values: {apr_data['apr'].tolist()}\n")
                 f.write(f"  APR timestamps: {[ts.strftime('%Y-%m-%d %H:%M:%S') if ts is not None else 'None' for ts in apr_data['timestamp']]}\n")
             
-            if not perf_data.empty:
-                f.write(f"  Performance values: {perf_data['apr'].tolist()}\n")
-                f.write(f"  Performance timestamps: {[ts.strftime('%Y-%m-%d %H:%M:%S') if ts is not None else 'None' for ts in perf_data['timestamp']]}\n")
-            
             f.write("\n")
     
     logger.info("Generated detailed graph data report")
@@ -540,8 +535,9 @@ def create_combined_time_series_graph(df):
     colors = px.colors.qualitative.Plotly[:len(unique_agents)]
     
     # IMPORTANT: Fixed y-axis range that always includes -100
-    min_apr = -110  # Fixed minimum to ensure -100 is visible
-    max_apr = 110   # Fixed maximum to ensure data is visible
+    # Since we're only showing positive values, adjust the range
+    min_apr = 0  # Start at 0
+    max_apr = max(df['apr'].max() * 1.1, 10)  # Add 10% padding, minimum of 10
     
     # Add background shapes for APR and Performance regions
     min_time = df['timestamp'].min()
@@ -557,16 +553,6 @@ def create_combined_time_series_graph(df):
         layer="below"
     )
     
-    # Add shape for Performance region (below zero)
-    fig.add_shape(
-        type="rect",
-        fillcolor="rgba(255, 230, 230, 0.3)",
-        line=dict(width=0),
-        y0=min_apr, y1=0,
-        x0=min_time, x1=max_time,
-        layer="below"
-    )
-    
     # Add zero line
     fig.add_shape(
         type="line",
@@ -575,7 +561,7 @@ def create_combined_time_series_graph(df):
         x0=min_time, x1=max_time
     )
     
-    # MODIFIED: Changed order of trace addition - add markers first, then lines
+    # MODIFIED: Changed order of trace addition - only need APR values now
     # Add data for each agent
     for i, agent_id in enumerate(unique_agents):
         agent_data = df[df['agent_id'] == agent_id].copy()
@@ -590,36 +576,6 @@ def create_combined_time_series_graph(df):
         for idx, row in agent_data.iterrows():
             logger.info(f"  Point {idx}: timestamp={row['timestamp']}, apr={row['apr']}, type={row['metric_type']}")
         
-        # First add scatter points for Performance values
-        perf_data = agent_data[agent_data['metric_type'] == 'Performance']
-        if not perf_data.empty:
-            logger.info(f"  Adding {len(perf_data)} Performance markers for {agent_name}")
-            for idx, row in perf_data.iterrows():
-                logger.info(f"    Performance marker: timestamp={row['timestamp']}, apr={row['apr']}")
-            
-                # Use explicit Python boolean for showlegend
-                is_first_point = bool(idx == perf_data.index[0])
-                fig.add_trace(
-                    go.Scatter(
-                        x=[row['timestamp']],
-                        y=[row['apr']],
-                        mode='markers',
-                        marker=dict(
-                            color='red',  # Force consistent color
-                            symbol='square',
-                            size=16,      # Make markers larger
-                            line=dict(
-                                width=2,
-                                color='black'
-                            )
-                        ),
-                        name=f'{agent_name} Perf',
-                        legendgroup=agent_name,
-                        showlegend=is_first_point,  # Use native Python boolean
-                        hovertemplate='Time: %{x}<br>Performance: %{y:.2f}<br>Agent: ' + agent_name + '<extra></extra>'
-                    )
-                )
-        
         # Now add scatter points for APR values
         apr_data = agent_data[agent_data['metric_type'] == 'APR']
         if not apr_data.empty:
@@ -665,7 +621,7 @@ def create_combined_time_series_graph(df):
     
     # Update layout - use simple boolean values everywhere
     fig.update_layout(
-        title="APR and Performance Values for All Agents",
+        title="APR Values for All Agents (Positive Values Only)",
         xaxis_title="Time",
         yaxis_title="Value",
         template="plotly_white",
@@ -690,8 +646,8 @@ def create_combined_time_series_graph(df):
         gridcolor='rgba(0,0,0,0.1)',
         range=[min_apr, max_apr],  # Fixed range
         tickmode='linear',
-        tick0=-100,
-        dtick=50
+        tick0=0,
+        dtick=10  # Adjusted for positive values
     )
     
     # Update x-axis
@@ -758,10 +714,10 @@ def create_combined_time_series_graph(df):
         
         # Simplified layout
         simple_fig.update_layout(
-            title="APR and Performance Values (Simplified View)",
+            title="APR Values (Simplified View - Positive Values Only)",
             xaxis_title="Time",
             yaxis_title="Value",
-            yaxis=dict(range=[-110, 110]),
+            yaxis=dict(range=[0, max_apr]),
             height=600,
             width=1000
         )
@@ -1179,78 +1135,7 @@ def create_visualizations():
     
     return fig_swaps_chain, fig_bridges_chain, fig_agents_registered, fig_tvl
 
-# Add new function to the bottom of the file, before the dashboard() function
-def add_diagnostic_controls(demo):
-    """Add diagnostic UI controls to help debug the difference between local and production"""
-    with gr.Column():
-        gr.Markdown("## Diagnostics")
-        
-        diagnostic_button = gr.Button("Run Data Diagnostics")
-        diagnostic_output = gr.Textbox(label="Diagnostic Results", lines=10)
-        
-        def run_diagnostics():
-            """Function to diagnose data issues"""
-            global global_df
-            
-            if global_df is None or global_df.empty:
-                return "No data available. Please click 'Refresh APR Data' first."
-            
-            # Gather diagnostics
-            result = []
-            result.append(f"=== DIAGNOSTIC REPORT ===")
-            result.append(f"Time: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
-            result.append(f"API Endpoint: {API_BASE_URL}")
-            result.append(f"Total data points: {len(global_df)}")
-            
-            unique_agents = global_df['agent_id'].unique()
-            result.append(f"Number of unique agents: {len(unique_agents)}")
-            
-            # Per-agent diagnostics
-            for agent_id in unique_agents:
-                agent_data = global_df[global_df['agent_id'] == agent_id]
-                agent_name = agent_data['agent_name'].iloc[0]
-                
-                result.append(f"\nAgent: {agent_name} (ID: {agent_id})")
-                result.append(f"  Data points: {len(agent_data)}")
-                
-                # Check APR values
-                apr_data = agent_data[agent_data['metric_type'] == 'APR']
-                perf_data = agent_data[agent_data['metric_type'] == 'Performance']
-                
-                result.append(f"  APR points: {len(apr_data)}")
-                if not apr_data.empty:
-                    result.append(f"  APR values: {apr_data['apr'].tolist()}")
-                
-                result.append(f"  Performance points: {len(perf_data)}")
-                if not perf_data.empty:
-                    result.append(f"  Performance values: {perf_data['apr'].tolist()}")
-            
-            # Write to file as well
-            with open("latest_diagnostics.txt", "w") as f:
-                f.write("\n".join(result))
-            
-            return "\n".join(result)
-        
-        # Fix for Gradio interface - use event listeners properly
-        try:
-            # Different Gradio versions have different APIs
-            # Try the newer approach first
-            diagnostic_button.click(
-                fn=run_diagnostics,
-                inputs=None,
-                outputs=diagnostic_output
-            )
-        except TypeError:
-            # Fall back to original approach
-            diagnostic_button.click(
-                fn=run_diagnostics,
-                inputs=[],
-                outputs=[diagnostic_output]
-            )
-    
-    return demo
-
-# Modify dashboard function to include diagnostics
+# Modify dashboard function to remove the diagnostics section
 def dashboard():
     with gr.Blocks() as demo:
         gr.Markdown("# Valory APR Metrics")
@@ -1308,9 +1193,6 @@ def dashboard():
                     font=dict(size=15)
                 )
                 combined_graph.value = placeholder_fig
-                
-                # Add diagnostics section
-                demo = add_diagnostic_controls(demo)
         
     return demo