The nma_nodesplit and nma_nodesplit_df classes contains the results from
running a node-splitting model with the function nma().
Objects of class nma_nodesplit inherit from the stan_nma class,
and contain the results of fitting a single node-split model. They have one
additional component, nodesplit, which gives the comparison that was
node-split as a length 2 vector.
Objects of class nma_nodesplit_df are tibble data frames with one row
for each node-split comparison and columns:
trt1, trt2Treatments forming the comparison
modelA list column containing the results of each model as a
nma_nodesplit object
Optionally, there will be an additional row for the consistency model if
this was fitted (e.g. by get_nodesplits(., include_consistency = TRUE))
with trt1 and trt2 both NA.