Conversation
complete_graph (include/graph/generators/complete.hpp): - Generates all n*(n-1) ordered pairs (u,v), u != v - Output naturally sorted by source_id then target_id - O(n^2) edge count; size warning in header and user-guide erdos_renyi_gnm (include/graph/generators/gnm.hpp): - Fixed edge count variant of Erdos-Renyi: selects exactly m distinct edges uniformly at random via rejection sampling on the G(n,p) position encoding (pos -> (u,v), same scheme as erdos_renyi.hpp) - Clamps m to n*(n-1) when too large - Output sorted by source_id after sampling Both generators: - Follow the established pattern: edge_list<VId> return type, weight_dist enum, uint64_t seed, VId template parameter - Added to generators.hpp umbrella header Tests (tests/generators/test_generators.cpp): - New TEST_CASEs for complete_graph and erdos_renyi_gnm covering: count, no self-loops, ids in range, distinctness, sorted, determinism, K(n) full coverage, gnm clamping - uint64_t template test extended to include both new generators Docs: - docs/user-guide/generators.md: TOC entries, individual-include lines, complete_graph and erdos_renyi_gnm sections with parameter tables, weight_dist options, and usage examples - agents/bgl_migration_strategy.md: mark G(n,m) and K(n) done in §9 table and API block; trim Remaining Gaps; update executive summary bullet, Appendix B Done note, scorecard row (67%->83%), and overall coverage (~46%->~47%)
watts_strogatz (include/graph/generators/watts_strogatz.hpp): - Ring lattice with k nearest neighbours + probability-beta rewiring - Avoids self-loops and duplicate undirected pairs; bidirectional edges - beta=0 yields pure ring lattice (N*K directed edges) rmat (include/graph/generators/rmat.hpp): - Recursive-matrix (Graph500-style) with (a,b,c,d) quadrant probabilities - 2^scale vertices, descends scale levels per edge - Default params are standard Graph500 (0.57,0.19,0.19,0.05) - Self-loops and duplicate directed edges removed plod (include/graph/generators/plod.hpp): - Palmer-Steffan power-law out-degree model - Per-vertex credit drawn from power law, edges to random targets - No self-loops or duplicates All three follow the established pattern (edge_list<VId> return, weight_dist enum, uint64_t seed, VId template param) and are added to the generators.hpp umbrella header. Tests: new TEST_CASEs for each covering no-self-loops, id range, distinctness, sorted, determinism, and model-specific invariants (WS ring-lattice count, RMAT m bound). uint64_t template test extended. Docs: - docs/user-guide/generators.md: TOC, includes, and full sections with parameter tables and examples for all three - agents/bgl_migration_strategy.md: mark WS/R-MAT/PLOD done in §9; only SSCA#2 remains; update executive summary, roadmap, Done note, and scorecard (generators 83%->95%, overall ~47%->~48%)
ssca (include/graph/generators/ssca.hpp): - Partitions vertices into consecutive cliques of random size [1, max_clique_size] - Dense intra-clique edges with up to max_parallel_edges parallel edges per pair (a defining SSCA#2 trait) - Sparse inter-clique edges, accept probability decays as 2^-(clique distance) - No self-loops; output sorted by source_id Completes full BGL graph-generator parity. Follows the established pattern (edge_list<VId> return, weight_dist enum, uint64_t seed, VId template param); added to generators.hpp umbrella header. Tests: new TEST_CASE covering no-self-loops, id range, sorted, determinism, and the empty-graph edge case (clique size 1, no inter-clique edges). uint64_t template test extended. Docs: - docs/user-guide/generators.md: TOC, include, full ssca section - agents/bgl_migration_strategy.md: mark SSCA#2 done; generators now 100% (full parity); move generators bullet from gaps to strengths in executive summary; clear Remaining Gaps; update API block, roadmap, and Done note
dimacs (include/graph/io/dimacs.hpp): - write_dimacs(os, g, problem="sp") — generic arc list, 1-indexed endpoints, edge weight auto-serialized when EV is formattable - write_dimacs_max_flow(os, g, source, sink, capacity_fn) — emits p/n/a lines - read_dimacs(is) -> dimacs_graph — handles c/p/n/a/e line kinds; normalizes 1-indexed endpoints to 0-indexed ids; parses max-flow, sp, and edge formats metis (include/graph/io/metis.hpp): - write_metis(os, g, with_weights=false) — treats graph as undirected; builds symmetric deduplicated adjacency; sets fmt=001 when with_weights=true and EV is formattable - read_metis(is) -> metis_graph — handles optional fmt 3-digit flag (vertex sizes / vertex weights / edge weights) and ncon field; normalizes to 0-indexed adjacency_list_text (include/graph/io/adjacency_list_text.hpp): - write_adjacency_list_text(os, g) — one line per vertex: "<id>: <t0> <t1> ..." - read_adjacency_list_text(is) -> adjacency_list_text_graph — accepts colon separator or whitespace-only; preserves vertex set (empty-adjacency lines) include/graph/io.hpp: extended umbrella to include all six I/O headers. Tests (tests/io/test_io.cpp): 15 new test cases covering write, read, roundtrip, and edge cases for all three formats (167 assertions in 34 test cases, all pass). Docs: - docs/user-guide/io.md: updated overview table, headers section, added full sections (API signatures, examples, reader return fields) for DIMACS, METIS, and adjacency-list-text - agents/bgl_migration_strategy.md: mark all three formats done in §8 table; remove "No DIMACS or METIS I/O" from gaps; move I/O to strengths (full BGL parity + JSON); strike roadmap items; Graph I/O scorecard 60% -> 100%; overall coverage ~48% -> ~50%
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
1 participant
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.
No description provided.