Package index • monecascale

Scalable Clustering Backends

Alternative backends producing moneca-class output

moneca_sbm(): Hierarchical DC-SBM Backend for MONECA Segmentation
moneca_sbm_install_graphtool(): Install the graph-tool Python Backend for moneca_sbm()

Flow Clustering

Infomap / Map Equation on mobility counts. Flat Infomap via igraph, with a recursive wrapper that synthesises a 2-3 level hierarchy. Output is a plain moneca-class object with a $flow_diagnostics slot.

moneca_flow(): Flow-based (Infomap / Map Equation) Backend for MONECA Segmentation

Bipartite Segmentation

Rectangular input (person x employer, worker x occupation). Returns two linked moneca-class objects - one row-view, one col-view - each consumable by moneca’s segment.() and plot_moneca_() stack.

moneca_bipartite(): Bipartite (Rectangular) DC-LBM Backend for MONECA Segmentation
format(<moneca_bipartite>): Format a `moneca_bipartite` object
print(<moneca_bipartite>): Print a `moneca_bipartite` object
summary(<moneca_bipartite>): Summary of a `moneca_bipartite` object
print(<summary.moneca_bipartite>): Print a `summary.moneca_bipartite` object

Auto-level Selection

Post-hoc selection of a hierarchy level from a moneca_sbm or moneca_bipartite fit via MDL elbow or level-to-level MI plateau. Composes with the segment.levels = “auto” wrapper on both backends.

auto_segment_levels(): Auto-select a Hierarchy Level for a `moneca` Fit
print(<auto_segment_levels>): Print an `auto_segment_levels` Result
format(<auto_segment_levels>): Format an `auto_segment_levels` Result

Out-of-core RR

Build a sparse relative-risk matrix from an out-of-core edge source (DuckDB / Parquet / CSV) without densifying to an R matrix. Output composes with the sparse-aware clustering backends.

rr_from_duckdb(): Build a Sparse Relative-Risk Matrix Out-of-Core via DuckDB