Download
GH Archive collects public GitHub activity and preserves it as a historical dataset for analysis. It polls GitHub’s event feed and groups the raw responses into compressed hourly JSON archives. The records cover actions such as commits, forks, issue activity, comments, and repository membership changes. Researchers can download specific hours, days, or longer periods with ordinary HTTP tools and process the events in their own systems. The dataset is also published through Google BigQuery for large-scale SQL-style exploration without downloading every archive. Its structure supports trend analysis, visualizations, machine learning, and open-source ecosystem research. The repository contains the crawler, supporting scripts, and website code, while the actual event files are hosted separately.
Features
- Hourly compressed JSON event archives
- Public GitHub activity captured from event feeds
- Direct HTTP downloads by hour, day, or date range
- Public BigQuery dataset for SQL-style analysis
- Raw event payloads preserved with minimal processing
- Crawler, scheduling, and data-processing utilities
Project Samples
