Allow user to pass custom contrasts for emmeans

[bzkrouse]

What's the feature?

For ard_emmeans_contrast(), allow more custom contrast statements that are possible in emmeans, like so:

warp.lm <- lm(breaks ~ wool*tension, data = warpbreaks)
warp.emm <- emmeans(warp.lm, ~ tension | wool)
contrast(warp.emm, list(mid.vs.ends = c(-1,2,-1)/2, lo.vs.hi = c(1,0,-1)))

[ddsjoberg]

Sounds good to me! Do you want to investigate and propose an API update?

[chizapoth]

I'm also having a similar question (not urgent at all):

I'm making AOVT01 (ancova). I tried to reproduce the output from tern (summarize_ancova), and the implementation that made the exact numbers to the table was emmeans::contrast(method = 'trt.vs.ctrl', adjust = 'none'). In ard_emmeans_contrast it is fixed to be emmeans::contrast(method = 'pairwise').

Would it be possible if at one point we allow users to select what methods to choose? I do understand that we don't want to give TOO MUCH flexibility.