Code Changes - Reranking Implementation

app.py

@@ -4,6 +4,7 @@ import json
 import logging
 
 from datetime import datetime
+import numpy as np
 
 
 # ==== Metrics Calculation
@@ -197,7 +198,7 @@ def plot_subset_metrics_old(subset_name):
     ax.set_xticklabels(keys, rotation=21, ha='right')
     return fig
 
-def plot_chunking_strategies(subset_name):
+def plot_chunking_strategies_old(subset_name):
     chunking_data = ragbench_details[subset_name]["chunking"]
 
     plt.figure(figsize=(8, 4))
@@ -222,6 +223,69 @@ def plot_chunking_strategies(subset_name):
     # Return plot as figure (Gradio accepts it)
     return plt.gcf()
 
+def plot_chunking_strategies(subset_name):
+    """Visualize chunking strategy distribution with enhanced formatting."""
+    try:
+        chunking_data = ragbench_details[subset_name]["chunking"]
+        
+        # Create figure with constrained layout to prevent clipping
+        fig, ax = plt.subplots(figsize=(10, 5), constrained_layout=True)
+        
+        # Prepare data
+        strategies = list(chunking_data.keys())
+        counts = list(chunking_data.values())
+        
+        # Create color gradient based on count values
+        colors = plt.cm.Blues(np.linspace(0.4, 1, len(strategies)))
+        
+        # Plot bars with different colors
+        bars = ax.bar(strategies, counts, color=colors, edgecolor='white', linewidth=0.7)
+        
+        # Add value labels with better positioning
+        for bar in bars:
+            height = bar.get_height()
+            ax.text(
+                bar.get_x() + bar.get_width()/2, 
+                height + max(counts)*0.02,  # Dynamic padding
+                f'{int(height):,}',  # Format with thousands separator
+                ha='center', 
+                va='bottom',
+                fontsize=10,
+                fontweight='bold'
+            )
+        
+        # Customize plot
+        ax.set_title(
+            f"Chunking Strategy Distribution\n{subset_name}",
+            fontsize=12,
+            pad=20,
+            fontweight='bold'
+        )
+        ax.set_ylabel("Number of Chunks", fontsize=10)
+        ax.set_xlabel("Chunking Method", fontsize=10, labelpad=10)
+        
+        # Rotate x-labels and adjust appearance
+        ax.set_xticks(range(len(strategies)))
+        ax.set_xticklabels(
+            strategies, 
+            rotation=30, 
+            ha='right',
+            fontsize=9,
+            rotation_mode='anchor'
+        )
+        
+        # Improve grid and spines
+        ax.grid(axis='y', linestyle=':', alpha=0.6)
+        ax.spines[['top', 'right']].set_visible(False)
+        
+        # Auto-scale y-axis with 10% headroom
+        ax.set_ylim(0, max(counts) * 1.1)
+        
+        return fig
+        
+    except Exception as e:
+        print(f"Error plotting chunking strategies: {str(e)}")
+
 # Initialize with first subset's summary
 initial_subset = available_subsets[0] if available_subsets else None
 initial_plot = plot_subset_metrics(initial_subset) if initial_subset else "No data available"