API Reference

Complete API documentation for backbone functions across 14 modules (the hypergraph page covers both hypergraph and hypergraph_io). API examples are standardized on nx.les_miserables_graph() for non-bipartite methods and nx.davis_southern_women_graph() for bipartite methods. Each module page includes Sphinx Gallery links showing score-then-filter visualizations for the listed functions. Each function entry includes complexity information (time/space), with an aggregate summary in Complexity Summary.

Module	Functions	Description
Statistical Methods	9	Hypothesis-testing methods plus short alias names
Structural Methods	14	Topology-based methods (threshold, sparsification, linkage, centrality, spanning tree, salience, metric, planarity)
Proximity Methods	12	Neighborhood-similarity edge scoring (Jaccard, Dice, cosine, Adamic-Adar)
Hybrid Methods	1	Combined statistical/structural methods (GLAB)
Bipartite Methods	16	Projection backbones, fixed null models, and high-level wrappers
Hypergraph Methods	23	MDL backbone, compression ratio, intersection / s-line graph, inclusion (toplex) reduction, order filter, s-components, and statistically validated hypergraphs / cores
Unweighted Sparsification	3	Sparsification for unweighted graphs (LSpar, local degree)
Filters	6	Post-hoc filtering utilities, multiple-testing correction, and graph-conversion support
Measures	7	Evaluation measures for comparing backbones
Visualization	3	Graph comparison and plotting helpers for backbone differences
Describe	1	Methods-section text and citations for a backbone (describe_backbone)
Convert	1	NumPy/SciPy adjacency matrices and pandas edge lists to NetworkX graphs (to_graph)
Temporal	15	Temporal networks as snapshots, events, or spans: form detection, conversions, binning, pandas event tables, and temporal backbone scores (observation fraction, persistence, backbone stability, weight trend, hidden Markov model of edge activity)