Temporal Coupling
Identifies files that frequently change together in pull requests. This analysis helps reveal hidden dependencies and areas of the codebase that are tightly coupled.
Available dimensions:
- workload name
- start date
Feature Flag
This feature is currently behind a feature flag. To enable it, set FEATURE_TEMPORAL_COUPLING=true.
UI
The Temporal Coupling analysis provides both table and ribbon chart views of file pairs.
Table View
The results show pairs of files that co-occur in pull requests:
- File A / File B: The paths of the two files that changed together
- Co-changes: The number of pull requests where both files were modified together
- Percentage: The percentage of total analyzed PRs where this co-occurrence happened
High co-change counts or percentages often indicate that these files are logically coupled and might benefit from refactoring to improve modularity.
Ribbon View
The ribbon view displays file-to-file coupling links, where thicker ribbons represent higher co-change counts.