Skip to contents

Trial contribution weights from DA, AD, and SWADA

Source: R/constructions.R
compare_weights.Rd

da() generally uses different inverse-variance weights for each subgroup series. ad() uses the interaction-series weights. swada() applies one common vector to both subgroup series (swada_reference and swada_nonref are identical); under the default this equals the AD weights.

Usage

compare_weights(da, ad = NULL, swada = NULL)

Arguments

da

A da() or conventional() fit.

ad

An ad() fit, or a swada() fit in the historical two-argument call.

swada

A swada() fit. Optional when ad is already a SWADA object.

Value

A tibble with one row per trial and class interactmeta_compare_weights (plottable).

Details

Warns if the supplied fits do not share the same engine and method.

Historical two-argument call compare_weights(conventional, swada) still works. CAMS information fractions are not SWADA weights; compare pooled IF with compare_ifrac().