A Linear-Time Alternative To t-SNE for Dimensionality Reduction and Fast Visualisation
Abstract
This release modernises the project structure and development workflow to align with current Python packaging standards. Key Changes: Infrastructure: Migrated build backend to hatchling (PEP 517) and added pyproject.toml dependency groups. Reliability: Introduced a comprehensive test suite with golden output regression tests for 2D/3D landmark modes (90%+ coverage). Automation: Integrated ruff for linting/formatting and established GitHub Actions for automated testing and Trusted Publishing. Tooling: Added a built-in diagnostic utility (lt-check) to verify installations and C++ build tool requirements. Modernisation: Raised minimum Python version to 3.9 to leverage PEP 604 type hints and updated the internal API to follow scikit-learn attribute conventions.
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 15 February 2026