Covariance edges of a distributional SEM (residual rho12 and higher-level corpair)

Returns the covariance allowances declared via covary() — double-headed arcs that are deliberately kept separate from paths() (which stays directed-only, including any x -> rho12 directed path into the correlation component). Residual (rho12, within-observation) and higher-level (corpair, between-unit random-effect) edges are reported together with a class column that distinguishes them; they answer different biological questions and are never collapsed. A declared edge also makes basis_set() / dsep() drop the y1 _||_ y2 independence claim.

Usage

covariances(object, ...)

# S3 method for class 'drm_sem'
covariances(object, ...)

Arguments

object

A drm_sem object.

...

Unused.

Value

A drm_covariances data frame with columns y1, y2, class ("residual" / "higher_level"), level, structure, label.

References

Shipley B (2016). Cause and Correlation in Biology: A User's Guide to Path Analysis, Structural Equations and Causal Inference with R, 2nd edition. Cambridge University Press, Cambridge.

Bollen KA (1989). Structural Equations with Latent Variables. Wiley, New York.

See also

covary(), paths(), basis_set().

Examples

if (FALSE) { # \dontrun{
sem <- drm_sem(
  activity = drm_node(drmTMB::bf(activity ~ x), family = stats::gaussian()),
  boldness = drm_node(drmTMB::bf(boldness ~ x), family = stats::gaussian()),
  data = dat,
  covariances = covary("activity", "boldness"))
covariances(sem)   # the residual rho12 edge, reported separately from paths()
} # }