v0.3.1: Expanded Python support & cross platform CI

This release focuses on expanding platform compatibility, robustifying the testing
infrastructure, and improving documentation.

• Enhancements

  • Python 3.14 Support: Added support for Python 3.14 in the build configuration.
  • Broadened Python Support: Fully tested support for Python 3.11, 3.12, 3.13, and 3.14.
  • Cross-Platform CI:
    • Added dedicated CI workflows for macOS and Windows to ensure native compatibility.
    • Added a Conda CI workflow to verify installation and testing using conda-forge dependencies.

Full Changelog: v0.3.0...v0.3.1

v0.3.0: Architecture Update

This release transforms dsm2dtm into a robust, developer-friendly library with rigorous testing.

• Major Features & API Changes

  • Pure Python: Dropped dependencies on SAGA/GDAL CLI tools. Now fully installable via pip on any platform.
  • Significant Speedups: Optimized core algorithms for faster processing.
  • New API: generate_dtm(path_or_object) now returns (dtm_array, profile) for flexible integration. Use save_dtm() to write to disk.
  • Flexible Input: Support for passing open rasterio objects directly to the pipeline.

• Robustness

  • Smart Reprojection: Now uses pyproj for accurate global UTM zone estimation.
  • Quality Assurance: New monotonicity checks ensure the generated DTM doesn't accidentally float above the input surface.

• Testing & CI

  • Zero Bloat Tests: Unit tests now use synthetic data on-the-fly.
  • Real World Verification: Integration tests automatically fetch valid datasets to verify accuracy.
  • Stress Testing: Added suite for large DSMs, cliffs, noise, and sparse data.
  • Automated Publishing: New workflows for PyPI and Conda Forge updates.

Full Changelog: v0.2.0...v0.3.0

Test data

What's Changed

• this release just contain test files to be used by the CI pipeline.

v0.2

Initial release