Customize categorical term labels with a glue pattern about broom.helpers HOT 6 CLOSED

Note: should we consider two additional tidyselectors: all_treatment_contrasts() and all_sum_contrasts()?

from broom.helpers.

A simpler solution:

• tidy_add_contrasts() all add a column reference_term
• tidy_add_term_labels() populate a column reference_label
• new selectors all_treatment_coded() and all_sum_coded() to select categorical variables with sum or treatment contrasts
• new function tidy_rename_term_labels() with a glue pattern based on other cols

from broom.helpers.

Bonus question: do we need a selector all_intercepts()?

from broom.helpers.

After thinking, I'm feeling that a better approach would be to implement a glue pattern for categorical terms through model_list_terms_levels() and tidy_add_term_lables()

Such an approach will allow to take into account the term labels when computing interaction terms labels.

from broom.helpers.

Note: should we consider two additional tidyselectors: all_treatment_contrasts() and all_sum_contrasts()?

@larmarange question for you: At the moment in the PR I am working on, I made the following contrast functions: all_treatment_contrasts(), all_sum_contrasts(), all_poly_contrasts(), all_helmert_contrasts().

Would you prefer a single function with an argument to select the type, e.g. all_contrasts(type = c("treatment", "sum", "poly", "helmert")) ? I think I prefer one contrast function, but you should decide ☺️

from broom.helpers.

one contrast function is great.

I will review the PR ASAP (not sure today, but will try tomorrow)

from broom.helpers.