pkg::fun doesn't work in dplyr::funs

[ghost]

This works:

slci_msa_wide <- slci_msa %>%
  dplyr::select(principal_city_property, placement_care_type, value_msa) %>%
  tidyr::spread(placement_care_type, value_msa) %>%
  dplyr::left_join(
    dplyr::distinct(slci_msa, principal_city_property, pop_rank)
  ) %>%
  dplyr::arrange(-A) %>%
  dplyr::mutate(cost_rank = row_number()) %>%
  dplyr::arrange(A) %>%
  dplyr::mutate(affordability_rank = row_number()) %>%
  dplyr::select(principal_city_property, dplyr::contains("rank"), A, R, Z) %>%
  dplyr::mutate_at(dplyr::vars(A, R, Z), dplyr::funs(round)) %>%
  dplyr::mutate_at(dplyr::vars(A, R, Z), dplyr::funs(dollar))

But this doesn't, and the only think I changed is the last line, resulting in an error: "Error: cannot coerce type 'closure' to vector of type 'character'"

slci_msa_wide <- slci_msa %>%
  dplyr::select(principal_city_property, placement_care_type, value_msa) %>%
  tidyr::spread(placement_care_type, value_msa) %>%
  dplyr::left_join(
    dplyr::distinct(slci_msa, principal_city_property, pop_rank)
  ) %>%
  dplyr::arrange(-A) %>%
  dplyr::mutate(cost_rank = row_number()) %>%
  dplyr::arrange(A) %>%
  dplyr::mutate(affordability_rank = row_number()) %>%
  dplyr::select(principal_city_property, dplyr::contains("rank"), A, R, Z) %>%
  dplyr::mutate_at(dplyr::vars(A, R, Z), dplyr::funs(round)) %>%
  dplyr::mutate_at(dplyr::vars(A, R, Z), dplyr::funs(scales::dollar))

I think I understand why this is happening, but it causes me to violate best practice in calling functions, which is to call out the package, too.

[VincentGuyader]

I also find this "bug" :

iris %>%
  mutate(NEW=plyr::revalue(Species,c("virginica"="a","setosa"="b"))) %>%
  summary() # OK

iris %>%
  mutate_each(funs(plyr::revalue(.,c("virginica"="a","setosa"="b"))),Species) %>%
  summary() # DONT WORK.. but it should be, isnt'it ?

plop <-plyr::revalue # ninja tips
iris %>%
  mutate_each(funs(plop(.,c("virginica"="a","setosa"="b"))),Species) %>%
  summary() # OK

[krlmlr]

#2140 looks like a more recent duplicate, and #2057 like an older one.