Open edelarua opened 21 hours ago
For this function, perhaps we can look into a helper that will solve a more general problem: repeating analyses of two groups across multiple levels of a variable? For example, if we need to do a t-test for A vs B, and A vs C?
Where should this function live? It's a utility, and we keep our utility functions in cards (ie functions for helping us create ARDs), or in cardx where we do analysis.
jack +1 cards daniel +1 don't care, but leaning to cards
Evaluate possibility of adding a method to summarize Cox PH pairwise models like
tern::s_coxph_pairwise()
, where each group is compared in isolation to the reference group when constructing the Cox regression model(s) to be summarized.Can currently be done using
ard_regression()
, but each pairwise model has to be calculated and run through the function separately and then the results combined.See https://insightsengineering.github.io/tern/latest-tag/reference/survival_coxph_pairwise.html for details on the tern function.