For Bayesian CAMS (engine = "bayesmeta"), returns the list of stored
univariate fits: $overall (bayesmeta::bmr() of \(m_j\) on \(\pi_j\)),
$interaction (bayesmeta::bayesmeta() of \(g_j\)), and $trial_blocks.
Those objects are the bayesmeta / DIRECT backend, not CAMS estimands.
Plot them with the native bayesmeta methods
(vignette("cams")). For frequentist CAMS the
analogue is $metafor.
Arguments
- fit
A
cams()/da()/ad()/swada()/conventional()object.
Value
For Bayesian CAMS, the list with $overall, $interaction, and
$trial_blocks. For frequentist CAMS, fit$metafor. For DA / AD /
SWADA / conventional, the stored $fits list. NULL if none is stored.