Random effects with explicit grouping columns

Builds the structure of default_random_effects() from coded data with grouping columns that can be set explicitly; each NULL falls back to the column tier_vars() detects.

Usage

build_default_random_effects(
  model_data,
  case_var = NULL,
  nest_var = NULL,
  study_var = NULL
)

Arguments

model_data

Coded data from process_scd_phases().

case_var, nest_var, study_var

Optional column names for the case-level grouping, the within-case grouping (e.g., behavior, skill, task), and the study-level grouping.

Value

A character string representing the random effects formula, or NULL when no case variable is found or the data hold a single series.

Examples

d <- data.frame(
  case = rep(c("C1", "C2"), each = 10),
  phase = rep(c("A", "B"), 10),
  level_AB = rep(c(0, 1), 10),
  outcome = rnorm(20)
)
build_default_random_effects(d)
#> [1] "(1 + level_AB | case)"