Conversation
Graphify-Labs#3859 items 3 and 5 (items 1, 2, 4 are covered by Graphify-Labs#3926): - Concepts table: note the silent networkx Louvain fallback when the optional `leiden` extra is not installed, instead of presenting Leiden as the only algorithm. - Extras table `leiden` row: document the fallback and that switching algorithms reshuffles every community, so a full `graphify label` run is needed for fresh names. - New 'Label reuse after re-clustering' callout documenting .graphify_labels.json.sig: labels survive a re-cluster only while a community's membership signature is unchanged; changed/new communities get hub-derived placeholders until a full `graphify label` run. Facts verified against graphify/cli.py (labels + .sig write/reuse), graphify/cluster.py::_partition and pyproject.toml.
|
Thanks for the pull request, @ffffff55. A maintainer will review it soon. Want to talk it through while it is in review? Come join us on our Discord server. For longer-form discussion there is also GitHub Discussions. A couple of things that speed up review: make sure the test suite passes on Python 3.10 and 3.13, and that the change keeps extraction deterministic. |
![graphify-labs[bot]](https://avatars.githubusercontent.com/in/4356270?s=60&v=4){kind=small}
There was a problem hiding this comment.
Graphify reviewed this change.
Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).
Graphify review — findings
Documents that community detection falls back to networkx Louvain when the optional leiden extra isn't installed, and that switching backends later reshuffles every community. Also explains that saved labels in .graphify_labels.json are reused on cluster-only/update only when a community's membership signature is unchanged. Changed or new communities get hub-derived placeholder names until a full graphify label run.
No blocking issues surfaced.
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 24 functions depend on the 24 functions this change touches.
Health — grade A; no new coupling hotspots.
Verification — 24 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 24 function(s) in the blast radius were not formally verified this run
Test selection
Test selection
308 of 308 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/test_affected_cli.py— full-run-safetytests/test_affected_member_seed.py— full-run-safetytests/test_agents_platform.py— full-run-safetytests/test_analyze.py— full-run-safetytests/test_anthropic_custom_endpoint.py— full-run-safetytests/test_antigravity_install.py— full-run-safetytests/test_apm_fallback_version.py— full-run-safetytests/test_architecture_doc.py— full-run-safetytests/test_astro_extraction.py— full-run-safetytests/test_astro_import_ids.py— full-run-safetytests/test_atomic_canvas_export.py— full-run-safetytests/test_atomic_version_stamp.py— full-run-safetytests/test_atomic_writes.py— full-run-safetytests/test_backend_env_isolation.py— full-run-safetytests/test_backend_extras.py— full-run-safetytests/test_benchmark.py— full-run-safetytests/test_benchmark_raw_graph.py— full-run-safetytests/test_blade_extractor.py— full-run-safetytests/test_build.py— full-run-safetytests/test_build_merge_dedup_scope.py— full-run-safetytests/test_build_merge_hyperedges_and_prune.py— full-run-safetytests/test_build_merge_shrink_guard.py— full-run-safetytests/test_builtin_global_type_refs.py— full-run-safetytests/test_cache.py— full-run-safetytests/test_callflow_html.py— full-run-safetytests/test_cargo_introspect.py— full-run-safetytests/test_cargo_missing_manifest.py— full-run-safetytests/test_carried_hyperedge_remap.py— full-run-safetytests/test_case_sensitive_resolution.py— full-run-safetytests/test_charmap_encoding.py— full-run-safetytests/test_chunking.py— full-run-safetytests/test_cjs_module_extension.py— full-run-safetytests/test_claude_cli_backend.py— full-run-safetytests/test_claude_md.py— full-run-safetytests/test_cli_broken_pipe.py— full-run-safetytests/test_cli_export.py— full-run-safetytests/test_cli_help.py— full-run-safetytests/test_cluster.py— full-run-safetytests/test_cobol_extractor.py— full-run-safetytests/test_codebuddy.py— full-run-safetytests/test_community_hub_labels.py— full-run-safetytests/test_community_labels_skill.py— full-run-safetytests/test_confidence.py— full-run-safetytests/test_corrupt_graph_json.py— full-run-safetytests/test_cpp_nested_and_cli.py— full-run-safetytests/test_cpp_objc_cross_file_calls.py— full-run-safetytests/test_cpp_preprocess.py— full-run-safetytests/test_cross_extension_reexport_self_cycle.py— full-run-safetytests/test_cross_language_call_resolution.py— full-run-safetytests/test_cross_repo_external_call_guards.py— full-run-safety- … and 258 more
non-code file(s) changed (
README.md) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
README.md) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
This branch has not been deployed
1 participant
Addresses items 3 and 5 of #3859 (items 1, 2, and 4 are already covered by #3926).
What
Three README additions, README-only (+4/−2):
leidenextra is not installed.leidenrow — documents the fallback and that switching algorithms reshuffles every community, so a fullgraphify labelrun is needed for fresh names..graphify_labels.json.sig: a saved label is reused only while its community's membership signature is unchanged; changed or new communities get hub-derived placeholder names untilgraphify labelruns again; after the community set changes (a different--resolution, or installing theleidenextra), a fullgraphify labelis the only way to refresh every name.Why
From #3859: the
.sigsidecar decides which labels survive a re-cluster but was not documented anywhere in the README (cf. #3808), and the README described Leiden as the clustering algorithm without mentioning the Louvain fallback — each cost the reporter a failed run or a wrong assumption.How checked
.graphify_labels.json+.sigwrite/reuse path ingraphify/cli.py(community_member_sigsreuse guard, hub-fill for changed communities), the graspologic → networkx Louvain fallback chain ingraphify/cluster.py::_partition, and theleidenextra inpyproject.toml.uv run pre-commit run --files README.md— passes (skillgen-check Passed; ruff has no files to check).