Use srtree from shapely instead of rtree package

[nreinicke]

Updates the internal RTree to be the SRTree from shapely rather than using the rtree package. This allows us to pickle and unpickle the matcher without corrupting the index.

Also adds support for the pixi build tool.

[nreinicke]

And this drops python 3.9 support and adds python 3.13 support

[Copilot]

Pull Request Overview

This PR replaces the rtree package dependency with Shapely's built-in STRtree to enable proper pickling/unpickling of matcher objects. It also adds pixi build tool support and updates Python version support (dropping 3.9, adding 3.13).

Key changes:

• Switched from rtree.index.Index to shapely.strtree.STRtree for spatial indexing
• Removed rtree dependency from pyproject.toml
• Added version constraints to all dependencies
• Configured pixi as an alternative build tool

Reviewed Changes

Copilot reviewed 9 out of 11 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
pyproject.toml	Removed rtree dependency, added version constraints, added pixi configuration, updated Python version support
mappymatch/maps/nx/nx_map.py	Replaced rtree with STRtree, updated nearest_road implementation
mappymatch/maps/igraph/igraph_map.py	Replaced rtree with STRtree, updated _nearest_edge_index implementation
mappymatch/matchers/lcss/lcss.py	Removed unused match_trace_batch method that had rtree serialization issues
.github/workflows/*.yml	Updated Python version matrix to 3.10-3.13
environment*.yml	Deleted conda environment files (replaced by pixi)

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[nreinicke]

This also introduces an optional parameter "additional_metadata_keys" into the NxMap.from_geofence method that will allow users to pass in additional keys they want to persist through the matching. For example, they could pass additional_metadata_keys=["highway"] and those OSM values would show up in the final match result.

[nreinicke]

@jhoshiko - this should be ready for your review!

[jhoshiko]

Ok, All of the changes looks good to me. I pulled the branch and ran all of the unit tests which are all passing and also put together a quick notebook to test the branch which also went flawlessly (all testing was python=3.13).

I also noticed your note about the memory usage while building the tree and I will also keep an eye out for memory issues in the future, but I think we can merge this in! Thanks for putting this together.