Explicitly specify dplyr::across, rather than use (magic) across

While `dplyr` appears to automagically provide its own `across` inside
`group_by`, we still want to explicitly use `dplyr::` to
- satisfy package checks,
- continue to work if `dplyr` removes this magic (e.g., if the maintainers don't
  like this magic ignoring any user-defined/attached non-`dplyr` `across`
  function), and
- be clear to code readers&editors where the function comes from.

Author: lcbrooks

R/get_test_data.R

@@ -51,7 +51,7 @@ get_test_data <- function(recipe, x){
         .fns = ~ !is.na(.x)
       )
     ) %>%
-    epiprocess::group_by(across(dplyr::all_of(groups))) %>%
+    epiprocess::group_by(dplyr::across(dplyr::all_of(groups))) %>%
     dplyr::slice_tail(n = max(max_lags) + 1) %>%
     epiprocess::ungroup()