Delete debug_analyzer.py

debug_analyzer.py

@@ -1,135 +0,0 @@
-"""
-Standalone script to analyze debug output and diagnose issues
-"""
-import json
-import numpy as np
-import matplotlib.pyplot as plt
-from collections import defaultdict
-from pathlib import Path
-
-def analyze_reid_debug(debug_file='reid_debug.json'):
-    """Analyze the debug output file and provide insights"""
-    
-    if not Path(debug_file).exists():
-        print(f"Debug file {debug_file} not found")
-        return
-        
-    with open(debug_file, 'r') as f:
-        data = json.load(f)
-        
-    print("="*80)
-    print("REID DEBUG ANALYSIS")
-    print("="*80)
-    
-    summary = data.get('summary', {})
-    log = data.get('log', [])
-    
-    # Overall stats
-    print("\n📊 OVERALL STATISTICS:")
-    print(f"  Total frames: {summary.get('total_frames', 0)}")
-    print(f"  Temp IDs created: {summary.get('temp_ids_created', 0)}")
-    print(f"  Active dogs: {summary.get('active_dogs', 0)}")
-    print(f"  Sleeping tracks: {summary.get('sleeping_tracks', 0)}")
-    
-    # Quality analysis
-    quality_stats = summary.get('quality_stats', {})
-    if quality_stats:
-        print("\n🎯 FEATURE QUALITY:")
-        print(f"  Average: {quality_stats.get('avg', 0):.3f}")
-        print(f"  Range: [{quality_stats.get('min', 0):.3f} - {quality_stats.get('max', 0):.3f}]")
-        
-        if quality_stats.get('avg', 0) < 0.5:
-            print("  ⚠️ LOW AVERAGE QUALITY - Consider:")
-            print("    • Better lighting")
-            print("    • Higher resolution")
-            print("    • Less motion blur")
-            
-    # Storage stats
-    storage_stats = summary.get('storage_stats', {})
-    if storage_stats:
-        print("\n💾 STORAGE EFFICIENCY:")
-        total_stored = sum(v for k, v in storage_stats.items() if 'stored' in k)
-        total_rejected = sum(v for k, v in storage_stats.items() if 'rejected' in k)
-        
-        if total_stored + total_rejected > 0:
-            rejection_rate = total_rejected / (total_stored + total_rejected) * 100
-            print(f"  Stored: {total_stored}")
-            print(f"  Rejected: {total_rejected}")
-            print(f"  Rejection rate: {rejection_rate:.1f}%")
-            
-            if rejection_rate > 50:
-                print("  ⚠️ HIGH REJECTION RATE - Many low-quality features")
-                
-    # Analyze adaptive thresholds
-    adaptive_entries = [e for e in log if e.get('operation') == 'adaptive_threshold']
-    if adaptive_entries:
-        print("\n🎯 ADAPTIVE THRESHOLD ANALYSIS:")
-        adjustments = [e['details']['adjustment'] for e in adaptive_entries]
-        confidence_levels = [e['details']['confidence_level'] for e in adaptive_entries]
-        
-        level_counts = defaultdict(int)
-        for level in confidence_levels:
-            level_counts[level] += 1
-            
-        print(f"  Total adaptations: {len(adaptive_entries)}")
-        for level, count in level_counts.items():
-            pct = count / len(adaptive_entries) * 100
-            print(f"  {level}: {count} ({pct:.1f}%)")
-            
-        avg_adjustment = np.mean(adjustments)
-        print(f"  Average adjustment: {avg_adjustment:.3f}")
-        
-    # Analyze match patterns
-    quality_entries = [e for e in log if e.get('operation') == 'quality_scoring']
-    if quality_entries:
-        print("\n📈 QUALITY PATTERNS:")
-        
-        # Extract component scores
-        components = defaultdict(list)
-        for entry in quality_entries:
-            for comp, score in entry['details'].get('components', {}).items():
-                components[comp].append(score)
-                
-        print("  Average component scores:")
-        for comp, scores in components.items():
-            print(f"    {comp}: {np.mean(scores):.3f}")
-            
-        # Find weakest component
-        avg_scores = {comp: np.mean(scores) for comp, scores in components.items()}
-        weakest = min(avg_scores.items(), key=lambda x: x[1])
-        print(f"\n  ⚠️ Weakest component: {weakest[0]} ({weakest[1]:.3f})")
-        
-        if weakest[0] == 'sharpness':
-            print("    → Motion blur issue, consider:")
-            print("      • Faster shutter speed")
-            print("      • Better camera stabilization")
-            print("      • Higher frame rate")
-        elif weakest[0] == 'bbox_size':
-            print("    → Dogs too small/far, consider:")
-            print("      • Better camera placement")
-            print("      • Optical zoom")
-            print("      • Multiple cameras")
-        elif weakest[0] == 'centrality':
-            print("    → Dogs often at frame edges, consider:")
-            print("      • Wider field of view")
-            print("      • Camera repositioning")
-            
-    # Recommendations
-    print("\n🚀 RECOMMENDATIONS:")
-    
-    temp_id_ratio = summary.get('temp_ids_created', 0) / max(summary.get('total_frames', 1) / 100, 1)
-    if temp_id_ratio > 2:
-        print("  ❌ Too many IDs created")
-        print("    • Increase ReID threshold")
-        print("    • Check feature quality")
-    elif temp_id_ratio > 1:
-        print("  ⚠️ Moderate fragmentation")
-        print("    • Fine-tune thresholds")
-        print("    • Verify DeepSORT parameters")
-    else:
-        print("  ✅ Good ID stability")
-        
-    print("\n" + "="*80)
-
-if __name__ == "__main__":
-    analyze_reid_debug()