Build a component-coded treatment-by-component matrix

Splits multi-component treatment labels (e.g. "A + B") on sep.comps and returns the binary treatment-by-component design matrix C, where C[t, c] = 1 if treatment t contains component c. The inactive component (e.g. placebo) is represented by an all-zero row, matching the convention of netmeta::netcomb().

Usage

build_C_matrix(treatments, sep.comps = "+", inactive = NULL)

Arguments

treatments

Character vector of (unique) treatment labels.

sep.comps

Single character separating components in a treatment label. Default "+".

inactive

Optional name of the inactive/reference treatment or component (mapped to a zero row / dropped as a column).

Value

A binary matrix with one row per treatment and one column per component (treatments as row names, components as column names).

Examples

build_C_matrix(c("A", "B", "A + B", "placebo"), inactive = "placebo")
#>         A B
#> A       1 0
#> B       0 1
#> A + B   1 1
#> placebo 0 0