Criterion-weighted average of standardized path coefficients

average() returns model-averaged standardized path coefficients across the candidate set, weighting each model's standardize()d paths() by the selected criterion's weight (CBIC by default, or CICc when compare(..., criterion = "CICc") was used). Coefficients are matched on from, to and component, so a path present in only some candidates is averaged over the weight of the models that contain it (conditional averaging). This mirrors phylopath's average().

Usage

average(object, ...)

# S3 method for class 'drm_model_set'
average(object, ...)

# S3 method for class 'drm_compare'
average(object, method = c("sd_x", "latent"), ...)

Arguments

object: A drm_compare (from compare()) or a drm_model_set.
...: Passed to compare() when object is an unfitted model set.
method: Standardization passed to standardize().

Value

A data frame of averaged standardized paths with columns from, to, component, std.estimate (weighted mean) and weight_sum (total weight of the models containing that path).

References

Burnham KP, Anderson DR (2002). Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach, 2nd edition. Springer, New York.

van der Bijl W (2018). “phylopath: Easy Phylogenetic Path Analysis in R.” PeerJ, 6, e4718. doi:10.7717/peerj.4718 .

Examples

if (FALSE) { # \dontrun{
average(cmp)
} # }