I'm trying to export my CAGEset to a dataset for DEseq2 with consensusClustersDESeq2() but I'm a bit confused by how the grouping for the design matrix should be supplied

From the package vignette

ce$group <- factor(c("a", "a", "b", "b"))
dds <- consensusClustersDESeq2(ce, ~group)

I'm guessing the grouping has to take into account the originally supplied files (before replicate merging and sample reordering) but it would be nice to have a confirmation before doing DGE with mixed up sample labels :)

Thanks!