Grouping columns that define one data series (tier)

Returns the grouping columns present, ordered from outermost to innermost: study (find_study_var()), cluster (find_cluster_var()), case (find_case_var()), and a within-case nest variable such as behavior or setting (find_nest_var()). Each combination of these columns is one tier: one AB series whose change points are coded at its own onset.

Usage

tier_vars(data, time_var = find_time_var(data))

Arguments

data: Data frame.
time_var: Time column used to decide whether a nest variable splits concurrent series; defaults to find_time_var(data).

Value

Character vector of column names (possibly empty).

Details

A within-case nest variable is kept when time points repeat within the study-cluster-case key, that is, when a case carries several concurrently measured series, or, for a column named like a nest (behavior, setting, skill, ...), when phase onsets differ across its levels within a case (a multiple-baseline design across behaviors whose sessions do not overlap in time). Behaviors or settings within a participant form an additional level between measurements and participants; cases are nested within studies, so a case identifier repeated in two studies names two cases.

Examples

d <- data.frame(study = rep(c("S1", "S2"), each = 4), case = "C1",
                time = rep(1:4, 2), outcome = rnorm(8))
tier_vars(d)
#> [1] "study" "case"