[Feature Request]: tern: ANCOVA extensions and summaries

[munoztd0]

Feature description

• File: R\summarize_ancova.R
• Functions:
  • h_ancova()
    • Origin: Forked from tern::h_ancova()
    • Purpose: Adds support for weights_emmeans and optional interaction_item
    • Exported: No (helper).
  • s_ancova_j()
    • Origin: Extension of tern:::s_ancova()
    • Purpose: Returns 3 extra statistics: lsmean_se, lsmean_ci, lsmean_diffci; allows specifying weights_emmeans; handles NULL .ref_group by returning NA diffs
    • Exported: Yes.
  • s_summarize_ancova_j() and a_summarize_ancova_j()
    • Origin: Combines tern::s_summary() with ANCOVA-based estimates using s_ancova_j()
    • Purpose: Provide unadjusted summary plus adjusted mean and difference via ANCOVA; a_... is the formatted afun that supports ref_path and .spl_context
    • Exported: Yes (both)

[munoztd0]

Regarding #52 I created this sub-issue accordingly, please have a quick look and tell

1. this should be brought a as tern PR
   OR
2. this should stay in junco
   OR
3. I don't personally know
   OR
4. this should be brought a as tern PR and I will take it as it my code

[danielinteractive]

Thanks @munoztd0!

• h_ancova()
  • support for weights_emmeans and optional interaction_item can be brought as a PR to tern
• s_ancova_j()
  • Returns 3 extra statistics: lsmean_se, lsmean_ci, lsmean_diffci; allows specifying weights_emmeans; handles NULL .ref_group by returning NA diffs - can all be brought as a PR to tern
• s_summarize_ancova_j() and a_summarize_ancova_j()
  • Origin: Combines tern::s_summary() with ANCOVA-based estimates using s_ancova_j()
  • Recommend to keep in junco because it is a custom combo function and we need the ref_path thing