fix: support tsam_xarray 0.6.5 by using cluster_counts

[FBumann]

Problem

tsam_xarray 0.6.5 renamed AggregationResult.cluster_weights to cluster_counts, keeping the old name as a deprecated property. Since our pytest config promotes warnings to errors (pyproject.toml:214), the resulting FutureWarning fails almost the entire clustering suite on any environment that resolves to 0.6.5 — and the full extra's current floor of >= 0.6.1, < 1 already permits it, so a fresh install hits this today.

tests/test_clustering on main:

	tsam_xarray 0.6.1	0.6.5
main	286 passed	254 failed, 32 passed
this PR	—	286 passed

Every one of the 254 failures is the same FutureWarning: AggregationResult.cluster_weights is deprecated; use cluster_counts instead.

Change

• transform_accessor.py — cluster_weights → cluster_counts in build_cluster_weights
• pyproject.toml — full: >= 0.6.1 → >= 0.6.5; dev: ==0.6.1 → ==0.6.5

cluster_counts does not exist before 0.6.5, so the floor has to move with the rename rather than merely allowing the new version. The two changes are coupled and can't land separately.

Bonus: expand() gets ~3.6x faster

0.6.5 also vectorises unsegmented disaggregation, which speeds up expand() with no changes on our side. Measured on a 3-year, 40-cluster, 84-variable system, min of 10 runs:

tsam_xarray	expand()
0.6.1	0.262 s
0.6.5	0.073 s

cluster() is unaffected (0.096 s vs 0.095 s, within noise). Users on the full extra already resolved to 0.6.5 and have this; the exact pin in dev was holding CI and development environments on the slow path.

Note for #748

This touches build_cluster_weights, which #748 (reapply #739) also rewrites. Whoever lands that should keep cluster_counts — the rename is independent of the dim-naming work and is required for 0.6.5 either way.

🤖 Generated with Claude Code

Summary by CodeRabbit

• Bug Fixes

  • Improved cluster weighting metadata when building reduced flow systems.

• Chores

  • Updated the optional full and development dependency requirements for improved compatibility.

[coderabbitai]

📝 Walkthrough

Walkthrough

The reduced flow-system builder now sources cluster weights from aggregation cluster counts. Optional and development dependency requirements are updated to tsam_xarray 0.6.5.

Changes

Reduced clustering weights

Layer / File(s)	Summary
Clustering dependency constraints pyproject.toml	The full optional dependency requires tsam_xarray ≥ 0.6.5, while the development dependency pins 0.6.5.
Reduced cluster weight construction flixopt/transform_accessor.py	build_cluster_weights() derives cluster_weight from agg_result.cluster_counts.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• flixOpt/flixopt#593: Introduced and refactored the reduced clustering builder and its weight logic.
• flixOpt/flixopt#654: Rewired reduced clustering construction around tsam_xarray aggregation outputs.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description is detailed, but it does not follow the required template sections for type, related issues, testing, or checklist.	Add the template sections: Type of Change, Related Issues, Testing, and Checklist, and fill them out explicitly.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly summarizes the main change: updating flixopt for tsam_xarray 0.6.5 and switching to cluster_counts.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch chore/tsam-xarray-0.6.5

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}