how to apply() a function only to rowSums() > 0

Question

any thoughts on how to apply a function to only row where there is at least 1 value more than 0?

I tried:

    batch4 =
  raw_mat[,all_conds == "LR140 AM"] %>%
  apply(., 1, FUN = function(x) ifelse(sum(x) > 0,  zCompositions::cmultRepl(., output = "p-counts", z.warning = 1), x))
Error in rowSums(x) : 'x' must be an array of at least two dimensions 

but it seems rowSums is only taking 1 element of the row at the time?

    > batch4 =
+   raw_mat[,all_conds == "LR140 AM"] %>%
+   rowwise() %>% 
+   mutate(across(where(rowSums(.) > 0), zCompositions::cmultRepl(., output = "p-counts", z.warning = 1)))
Error in `mutate()`:
! Problem while
  computing `..1 =
  across(...)`.
ℹ The error occurred in row
  0.
Caused by error in `zCompositions::cmultRepl()`:
! Row(s) containing all zeros/unobserved values were found (check it out using zPatterns).
Run `rlang::last_error()` to see where the error occurred.

So in this case it looks like the rowSums filter is not working properly, as cmultRepl gets also "only zeros" rows.