feat: add __repr__, __copy__, __deepcopy__ to core classes#693
Open
feat: add __repr__, __copy__, __deepcopy__ to core classes#693
Conversation
Patch v0.10.1
2 tasks
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## dev #693 +/- ##
==========================================
- Coverage 93.59% 93.58% -0.01%
==========================================
Files 66 66
Lines 5120 5131 +11
==========================================
+ Hits 4792 4802 +10
- Misses 328 329 +1 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
scipy's diags_array now warns (FutureWarning) when int64 input is silently cast to float64 output, and will change this behavior in a future release. This would silently alter numerical results. Ensure degree and weight arrays are explicitly float before passing to diags_array in both laplacian() and normalized_hypergraph_laplacian(). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Two changes to make exception usage consistent: 1. IDNotFound now inherits from (XGIException, KeyError) instead of just KeyError. This means all XGI exceptions can be caught uniformly with `except XGIException`, while preserving backwards compat for code that catches KeyError. 2. Standardize XGIError vs ValueError: use ValueError for pure input validation (bad argument values like "probability not in [0,1]"), reserve XGIError for domain-specific errors (disconnected graph, wrong network type, frozen network). Changes affect: - generators/uniform.py: all input validation raises - generators/lattice.py: invalid k value - generators/simple.py: invalid edge/core size - algorithms/assortativity.py: invalid kind argument Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add explicit __all__ lists to every subpackage (core, algorithms, communities, convert, drawing, dynamics, generators, linalg, readwrite, utils) and aggregate them in the top-level xgi/__init__.py. This locks down the public API surface to 169 names, preventing internal modules from leaking through wildcard imports. Also fixes tests that accessed xgi.utilities.dual_dict (an internal module path that was only reachable via leaked wildcard imports) to use the correct public path xgi.dual_dict. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Remove read_json/write_json from readwrite __all__ (deprecated) - Remove IDDict, Trie, crest_r from utils __all__ (internal) - Update test imports to use direct module paths Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…stances (#689) * feat: add seed parameter to unseeded stochastic functions Add seed parameter to h_eigenvector_centrality, degree_assortativity, simulate_kuramoto, and random_edge_shuffle for reproducibility. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: update doctest for pandas dtype='str' change Pandas now uses dtype='str' instead of dtype='object' for string Index columns. Compare as list to avoid version-dependent repr. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: make spectral clustering test deterministic and fix pandas doctest Seed numpy before eigsh in spectral_clustering so ARPACK produces consistent results. Seed random data in kmeans tests. Update pandas dtype doctest to compare as list. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: add floating point tolerance to Laplacian eigenvalue test eigvalsh can return eigenvalues like -1.3e-17 for a positive semi-definite matrix due to floating point arithmetic. Use -1e-12 tolerance instead of strict >= 0. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: make spectral clustering deterministic across platforms Pass explicit v0 to eigsh when seed is provided, making ARPACK initialization deterministic. Relax test assertion to check core community membership rather than exact partition, since boundary nodes can be assigned differently across LAPACK implementations. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * refactor: migrate from global random state to local RNG instances Replace all np.random.seed()/random.seed() + global function calls with np.random.default_rng(seed) and local Generator instances. This eliminates global state pollution, is thread-safe, and follows Scientific Python best practices. All functions now accept int | np.random.Generator | None for the seed parameter. Removed all stdlib random usage from xgi source. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: update doctest expected value in largest_connected_hypergraph The new RNG stream produces a different random hypergraph, changing the size of the largest connected component from 6 to 8. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: update tutorial seed for changed random stream The fast_random_hypergraph with seed=2 now produces a different number of dyads due to the RNG migration, causing a color array length mismatch in the multilayer drawing tutorial. Changed to seed=8 which produces 10 dyads matching the hardcoded color lists. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: address PR review comments on seed/RNG migration - Simplify rng.choice calls (remove unnecessary np.array conversion, use direct list selection for edge_list) - Standardize seed docstrings to "int, numpy.random.Generator, or None" for all functions using np.random.default_rng(seed) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * fix: simplify rng.choice to select directly from list Address review suggestion to pass the list directly to rng.choice instead of sampling indices. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Add __repr__ to Hypergraph, DiHypergraph (inherited by SimplicialComplex) showing a constructor-like representation. Add __copy__ and __deepcopy__ delegating to the existing .copy() method. Also fix rng.choice doctest in random_edge_shuffle. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
leotrs
force-pushed
the
dunder-method-parity
branch
from
872a155 to
50dfeb4
Compare
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
__repr__toHypergraphandDiHypergraph(inherited bySimplicialComplex) showing a constructor-like representation__copy__and__deepcopy__to both classes, delegating to the existing.copy()methodrng.choicedoctest inrandom_edge_shuffle(numpy scalars → Python types)Test plan
copy(),deepcopy(), andrepr()verified on Hypergraph, DiHypergraph, and SimplicialComplex🤖 Generated with Claude Code