[x] An rbind = FALSE default argument would return the current data structure. rbind = TRUE would pass that through do.call("rbind", x) to return a single data frame.
[x] Add this to calls within prediction() methods. That should improve performance by eliminating the for-loop.
(But, unfortunately, for calculating variances in margins::margins(), we still need to loop over the list.)
rbind = FALSEdefault argument would return the current data structure.rbind = TRUEwould pass that throughdo.call("rbind", x)to return a single data frame.prediction()methods. That should improve performance by eliminating the for-loop.(But, unfortunately, for calculating variances in
margins::margins(), we still need to loop over the list.)