Fix Pareto frontier in Evolution Over Time chart

- Make frontier strictly monotonically increasing (only shows improvements)
- Use step pattern: horizontal line until new best, then vertical jump
- Frontier line now exactly passes through the frontier points

Co-authored-by: openhands <openhands@all-hands.dev>

visualizations.py

@@ -185,17 +185,38 @@ def create_evolution_over_time_chart(df: pd.DataFrame) -> go.Figure:
     
     fig = go.Figure()
     
-    # Add trend line first (so it's behind the markers)
+    # Add Pareto frontier line (monotonically increasing best score over time)
     if len(plot_df) > 1:
-        fig.add_trace(go.Scatter(
-            x=plot_df['release_date'],
-            y=plot_df[score_col],
-            mode='lines',
-            line=dict(color='#FFE165', width=2, dash='dash'),  # primary yellow
-            name='Trend',
-            hoverinfo='skip',
-            showlegend=False
-        ))
+        # Compute Pareto frontier: for each date, track the max score seen so far
+        frontier_dates = []
+        frontier_scores = []
+        max_score_so_far = float('-inf')
+        
+        for _, row in plot_df.iterrows():
+            current_score = row[score_col]
+            current_date = row['release_date']
+            
+            if current_score > max_score_so_far:
+                # This point is on the Pareto frontier
+                if frontier_dates:
+                    # Add a horizontal line segment from previous frontier point to this date
+                    frontier_dates.append(current_date)
+                    frontier_scores.append(max_score_so_far)
+                # Add the new frontier point
+                frontier_dates.append(current_date)
+                frontier_scores.append(current_score)
+                max_score_so_far = current_score
+        
+        if frontier_dates:
+            fig.add_trace(go.Scatter(
+                x=frontier_dates,
+                y=frontier_scores,
+                mode='lines',
+                line=dict(color='#FFE165', width=2),  # primary yellow, solid line
+                name='Pareto Frontier',
+                hoverinfo='skip',
+                showlegend=False
+            ))
     
     # Calculate axis ranges for domain coordinate conversion
     min_date = plot_df['release_date'].min()