fix(tabpfn): fix categorical handling in tabpfn unsupervised

[alperkent-cmi]

Fixes #7

[Copilot]

Pull request overview

Fixes TabPFN categorical handling so explicit dataset schema roles override cardinality inference.

Changes:

• Patches TabPFN’s nested inference helper to preserve explicit categorical indices.
• Propagates schema-derived feature and target metadata through generation.
• Adds validation, regression tests, and a patch-version bump.

Reviewed changes

Copilot reviewed 7 out of 8 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
synthdata/data.py	Derives categorical target metadata from the schema.
synthdata/generation/tabpfn_backend.py	Enforces explicit TabPFN categorical roles and validates targets.
synthdata/generation/pipeline.py	Forwards schema metadata to TabPFN generators.
tests/unit/test_data.py	Tests categorical and continuous target metadata.
tests/unit/test_generation_pipeline.py	Tests schema wiring and early validation.
tests/unit/test_tabpfn_backend.py	Tests the compatibility patch and target rejection.
pyproject.toml	Bumps the package version to 0.7.3.
uv.lock	Synchronizes the package version in the lockfile.

---

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