make7.py

import os
import pandas as pd
import re
from collections import defaultdict

def analyze_fif_paths(root_dir="split_fifs"):
    """
    Extract .fif file paths using regex and provide detailed analysis including emotion types.
    """
    # Define emotion mapping with corrected IDs
    emotion_mapping = {
        # Neutral (ID: 0)
        'neutralVideo': {'id': 0, 'emotion': 'neutral'},
        
        # Excited (ID: 1)
        'ratatChaseScene': {'id': 1, 'emotion': 'excited'},
        'Kenmiles': {'id': 1, 'emotion': 'excited'},
        'NBA': {'id': 1, 'emotion': 'excited'},
        
        # Happy (ID: 2)
        'scooby': {'id': 2, 'emotion': 'happy'},

	# Happy (ID: 3)
        'motivationalAuthor': {'id': 3, 'emotion': 'motivated'},
        
        # Relaxed (ID: 4)
        'waterfall': {'id': 4, 'emotion': 'relaxed'},
        'asmr': {'id': 4, 'emotion': 'relaxed'},
        'meditation': {'id': 4, 'emotion': 'relaxed'},
        
        # Sad (ID: 5)
        'saddogs': {'id': 5, 'emotion': 'sad'},
        'sadbaby': {'id': 5, 'emotion': 'sad'},
        'ChampDeath': {'id': 5, 'emotion': 'sad'},
        'sadBaby1': {'id': 5, 'emotion': 'sad'},
        
        # Horror (ID: 6)
        'conjuring': {'id': 6, 'emotion': 'horror'},
        
        # Angry (ID: 7)
        'angrydogs': {'id': 7, 'emotion': 'angry'},
        'thepiano': {'id': 7, 'emotion': 'angry'},
        
        # Disgusted (ID: 8)
        'trainspotting': {'id': 8, 'emotion': 'disgusted'},
        
        # Utility files (no emotion)
        'label': {'id': -1, 'emotion': 'utility'},
        'rating': {'id': -1, 'emotion': 'utility'},
        'navon': {'id': -1, 'emotion': 'utility'}
    }
    
    # Dictionary to store analysis data
    analysis = {
        'total_files': 0,
        'directory_counts': defaultdict(int),
        'subject_counts': defaultdict(int),
        'epoch_counts': defaultdict(int),
        'video_type_counts': defaultdict(int),
        'emotion_counts': defaultdict(int),
        'emotion_id_counts': defaultdict(int)
    }
    
    # List to store file paths
    file_data = []
    
    # Regex pattern for .fif files
    pattern = r'(\d+|Zacker)-mapped_epoch_(\d+)_(\w+)raw_interval_(\d+)\.raw\.fif$'
    
    # Walk through the directory
    for dirpath, dirnames, filenames in os.walk(root_dir):
        for filename in filenames:
            match = re.match(pattern, filename)
            if match:
                full_path = os.path.join(dirpath, filename)
                
                # Extract components from filename
                subject_id = match.group(1)
                epoch_num = match.group(2)
                video_type = match.group(3)
                
                # Get emotion data if video type is in mapping
                emotion_info = emotion_mapping.get(video_type, {'id': -1, 'emotion': 'unknown'})
                emotion_id = emotion_info['id']
                emotion = emotion_info['emotion']
                
                # Update counts
                analysis['total_files'] += 1
                analysis['directory_counts'][dirpath] += 1
                analysis['subject_counts'][subject_id] += 1
                analysis['epoch_counts'][epoch_num] += 1
                analysis['video_type_counts'][video_type] += 1
                if emotion != 'utility':  # Only count actual emotions
                    analysis['emotion_counts'][emotion] += 1
                    analysis['emotion_id_counts'][emotion_id] += 1
                
                # Add to file data
                file_data.append({
                    'file_path': full_path,
                    'subject_id': subject_id,
                    'epoch': int(epoch_num),
                    'video_type': video_type,
                    'emotion_id': emotion_id,
                    'emotion': emotion
                })
    
    # Create DataFrame
    df = pd.DataFrame(file_data)
    
    # Sort the DataFrame
    df = df.sort_values(['subject_id', 'epoch', 'file_path'])
    
    # Save the full analysis to CSV
    output_file = 'fif_file_analysis4.csv'
    df.to_csv(output_file, index=False)
    
    # Create a filtered DataFrame with only emotion-related files
    emotion_df = df[df['emotion_id'] >= 0].copy()
    emotion_output_file = 'emotion_files.csv'
    emotion_df.to_csv(emotion_output_file, index=False)
    
    # Print the analysis
    print_analysis(analysis, df)
    
    return df, analysis

def print_analysis(analysis, df):
    """Print detailed analysis of the .fif files."""
    
    print("\n" + "="*50)
    print("FIF FILES ANALYSIS REPORT")
    print("="*50)
    
    # Overall Statistics
    print("\n1. OVERALL STATISTICS")
    print("-"*30)
    print(f"Total .fif files found: {analysis['total_files']}")
    print(f"Number of subjects: {len(analysis['subject_counts'])}")
    print(f"Number of directories: {len(analysis['directory_counts'])}")
    
    # Subject Breakdown
    print("\n2. FILES PER SUBJECT")
    print("-"*30)
    for subject, count in sorted(analysis['subject_counts'].items()):
        print(f"Subject {subject}: {count} files")
    
    # Directory Breakdown
    print("\n3. FILES PER DIRECTORY")
    print("-"*30)
    for directory, count in sorted(analysis['directory_counts'].items()):
        rel_path = os.path.relpath(directory, "splif_fifs")
        print(f"{rel_path}: {count} files")
    
    # Emotion Analysis
    print("\n4. FILES PER EMOTION")
    print("-"*30)
    # Define the exact order of emotions with their IDs
    ordered_emotions = [
        (0, 'neutral'),
        (1, 'excited'),
        (2, 'happy'),
        (3, 'relaxed'),
        (4, 'sad'),
        (5, 'horror'),
        (6, 'angry'),
        (7, 'disgusted')
    ]
    
    for emotion_id, emotion_name in ordered_emotions:
        if emotion_id in analysis['emotion_id_counts']:
            count = analysis['emotion_id_counts'][emotion_id]
            print(f"Emotion ID {emotion_id} {emotion_name}: {count} files")
    
    # Print utility files separately
    utility_count = len(df[df['emotion'] == 'utility'])
    if utility_count > 0:
        print(f"\nUtility files (rating/label/navon): {utility_count} files")
    
    # Video Type Analysis
    print("\n5. FILES PER VIDEO TYPE")
    print("-"*30)
    for video_type, count in sorted(analysis['video_type_counts'].items()):
        print(f"{video_type}: {count} files")
    
    # Epoch Analysis
    print("\n6. FILES PER EPOCH")
    print("-"*30)
    for epoch, count in sorted(analysis['epoch_counts'].items(), key=lambda x: int(x[0])):
        print(f"Epoch {epoch}: {count} files")
    
    # CSV File Information
    print("\n7. CSV FILE OUTPUTS")
    print("-"*30)
    print("1. Full analysis file: fif_file_analysis.csv")
    print("2. Emotion-only file: emotion_files.csv")
    print("\nColumns:")
    for col in df.columns:
        print(f"- {col}")

    # Additional Statistics
    print("\n8. ADDITIONAL STATISTICS")
    print("-"*30)
    if analysis['total_files'] > 0:
        avg_files_per_dir = analysis['total_files'] / len(analysis['directory_counts'])
        avg_files_per_subject = analysis['total_files'] / len(analysis['subject_counts'])
        print(f"Average files per directory: {avg_files_per_dir:.2f}")
        print(f"Average files per subject: {avg_files_per_subject:.2f}")
        
        # Calculate percentages of emotion files
        emotion_files = sum(analysis['emotion_id_counts'].values())
        utility_files = utility_count
        print(f"\nEmotion files: {emotion_files} ({(emotion_files/analysis['total_files']*100):.1f}%)")
        print(f"Utility files: {utility_files} ({(utility_files/analysis['total_files']*100):.1f}%)")

if __name__ == "__main__":
    # Run analysis
    df, analysis = analyze_fif_paths()