Repository Activity Measure

analysis.py

# pip install gitpython numpy tqdm
import git
from collections import defaultdict
from datetime import timedelta
import numpy as np
import argparse
from tqdm import tqdm

def analyze_repo(repo_path, branch='main'):
    repo = git.Repo(repo_path)
    commits = list(repo.iter_commits(branch))
    commits.sort(key=lambda x: x.committed_datetime)  # Sort commits by date
    
    hourly_changes = defaultdict(int)
    commit_times = []
    total_changes = 0
    
    print(f"Analyzing {len(commits)} commits on branch '{branch}'")

    changes_per_hour = []
    last_100_changes = []

    pbar = tqdm(commits, desc="Processing commits")
    for i, commit in enumerate(pbar):
        hour = commit.committed_datetime.replace(minute=0, second=0, microsecond=0)
        changes = commit.stats.total['lines']
        hourly_changes[hour] += changes
        total_changes += changes
        commit_times.append(commit.committed_datetime)
        
        changes_per_hour.append(changes)
        last_100_changes.append(changes)
        if len(last_100_changes) > 100:
            last_100_changes.pop(0)

        median_changes = np.median(last_100_changes)
        pbar.set_postfix({
            'Date': f"{commit.committed_datetime.strftime('%Y-%m-%d')}",
            'Median changes/100': f"{median_changes:.2f}"
        }, refresh=True)

    active_hours = len(hourly_changes)
    avg_changes_per_hour = total_changes / active_hours if active_hours > 0 else 0
    
    # Calculate percentiles of changes per active hour
    median_changes_per_hour = np.median(list(hourly_changes.values()))
    percentile_25 = np.percentile(list(hourly_changes.values()), 25)
    percentile_75 = np.percentile(list(hourly_changes.values()), 75)
    
    # Calculate percentiles of time between commits
    time_between_commits = []
    for i in range(1, len(commit_times)):
        time_diff = (commit_times[i] - commit_times[i-1]).total_seconds() / 3600  # in hours
        time_between_commits.append(time_diff)
    
    median_time_between_commits = np.median(time_between_commits)
    percentile_25_time = np.percentile(time_between_commits, 25)
    percentile_75_time = np.percentile(time_between_commits, 75)
    
    return {
        'total_commits': len(commits),
        'total_changes': total_changes,
        'active_hours': active_hours,
        'avg_changes_per_hour': avg_changes_per_hour,
        'median_changes_per_hour': median_changes_per_hour,
        'percentile_25_changes': percentile_25,
        'percentile_75_changes': percentile_75,
        'median_time_between_commits': median_time_between_commits,
        'percentile_25_time': percentile_25_time,
        'percentile_75_time': percentile_75_time
    }

def print_summary(stats):
    print("\nOverall Summary:")
    print(f"Total commits: {stats['total_commits']}")
    print(f"Total changes: {stats['total_changes']} lines")
    print(f"Active hours: {stats['active_hours']}")
    print(f"Average changes per active hour: {stats['avg_changes_per_hour']:.2f} lines")
    print(f"Changes per active hour:")
    print(f"  25th percentile: {stats['percentile_25_changes']:.2f} lines")
    print(f"  50th percentile (median): {stats['median_changes_per_hour']:.2f} lines")
    print(f"  75th percentile: {stats['percentile_75_changes']:.2f} lines")
    print(f"Time between commits:")
    print(f"  25th percentile: {stats['percentile_25_time']:.2f} hours")
    print(f"  50th percentile (median): {stats['median_time_between_commits']:.2f} hours")
    print(f"  75th percentile: {stats['percentile_75_time']:.2f} hours")

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Analyze Git repository commit patterns.")
    parser.add_argument("repo_path", help="Path to the Git repository")
    parser.add_argument("-b", "--branch", default="main", help="Branch to analyze (default: main)")
    args = parser.parse_args()

    print(f"Analyzing repository: {args.repo_path}")
    print(f"Branch: {args.branch}")
    
    stats = analyze_repo(args.repo_path, args.branch)
    print_summary(stats)