multiply_bipartite: Construct a co-occurrence network via two-mode multiplication

In bibnets: Importing, Constructing, and Exporting Bibliometric Networks

Construct a co-occurrence network via two-mode multiplication

Description

The unified engine for all bibliometric networks. Operates entirely in sparse representation — never allocates a dense n x n matrix.

Usage

multiply_bipartite(
  B,
  mode = "columns",
  similarity = "none",
  threshold = 0,
  top_n = NULL,
  self_loops = FALSE
)

Arguments

B	A sparse bipartite matrix (works x entities), already weighted by the counting method.
mode	Character. "columns" for column-mode co-occurrence (e.g., co-citation), "rows" for row-mode (e.g., coupling).
similarity	Character. Normalization method (see normalize()).
threshold	Numeric. Minimum edge weight to retain.
top_n	Integer or NULL. If specified, keep only the top n most frequent nodes and return all edges among them.

Value

A data frame with columns from, to, weight, count.

bibnets documentation built on June 19, 2026, 1:06 a.m.