Functions for probing the taxon co-occurrence network, used for checking if the bias estimated by the center() function is fully-determined.

cooccurrence_matrix(mat, all = TRUE)

cooccurrence_network(mat, all = TRUE)

cooccurrence_components(mat, all = TRUE, enframe = FALSE)

Arguments

mat

A compositional error matrix

all

Whether to include taxa that aren't in any multi-taxon samples

enframe

Whether to "enframe" the returned vector

Details

The edges are weighted by the number of samples the pair of taxa co-occur in and so provide some information about the precision of the pairwise bias estimate.

Functions

  • cooccurrence_matrix: Adjacency matrix of the co-occurrence network

  • cooccurrence_network: Co-occurrence network as an igraph object

  • cooccurrence_components: Connected components of the co-occurrence network