Covariance cliques of a distributional SEM

Detects complete covariance sub-graphs (cliques) among declared covary() edges in a distributional SEM, partitioned by covariance class ("residual" vs "higher_level") and grouping level. For \(K \ge 3\) responses with pairwise residual covariances, cliques identify complete joint covariance blocks whose within-clique independence claims are suppressed in basis_set().

Usage

covariance_cliques(object, min_size = 2L, ...)

# S3 method for class 'drm_sem'
covariance_cliques(object, min_size = 2L, ...)

# S3 method for class 'drm_covariances'
covariance_cliques(object, min_size = 2L, ...)

# S3 method for class 'data.frame'
covariance_cliques(object, min_size = 2L, ...)

# Default S3 method
covariance_cliques(object, min_size = 2L, ...)

Arguments

object

A drm_sem, drm_covariances, or declaration object.

min_size

Minimum clique size to report (default 2L).

...

Unused.

Value

A drm_covariance_cliques data frame with columns clique_id, class, level, size, nodes, pairs_count, complete.

See also

covary(), covary_clique(), covariances(), basis_set().

Examples

cvs <- covary_clique(c("activity", "boldness", "exploration"))
covariance_cliques(cvs)
#> <drmSEM covariance cliques: 1 clique (K >= 3: 1)>
#>  clique_id    class level size                           nodes pairs_count
#>          1 residual  <NA>    3 activity, boldness, exploration           3
#>  complete
#>      TRUE