Skip to contents

Symbolize a metafor rma.mv fit (multilevel / multivariate meta-analysis, v0.13.1)

Source: R/symbolize-metafor.R
symbolize.rma.mv.Rd

Builds a symbolized_model from an rma.mv fit. Covers the multi-tier random-effects structure (random = list(~ 1 | study, ~ 1 | id) or nested ~ 1 | district / study), optional R-matrix structured random effects (e.g., phylogenetic via R = list(phylo = phylo.cor)), and meta-regression moderators.

Usage

# S3 method for class 'rma.mv'
symbolize(fit, symbols = NULL, units = NULL, context = NULL, ...)

Arguments

fit

A fitted statistical model object.

symbols

Optional named character vector mapping variable names to user-supplied LaTeX symbols, e.g. c(body_mass = "W_i", temperature = "T_i").

units

Optional named character vector mapping variable names to units, e.g. c(body_mass = "g", temperature = "C").

context

Optional short character description of the model, e.g. "avian body-size location-scale model".

...

Reserved for method-specific extra arguments.

Value

A symbolized_model object.

Details

Each random-effect tier gets a row in variance_components with kind = "heterogeneity" (unstructured) or kind = "structured" (when an R-matrix is attached). Tiers with an R-matrix render as u_r ~ N(0, sigma^2_r * R_r) in the LaTeX; structured tiers also appear in sym$metadata$structured_random with their matrix dimension.