group_walk / group_map not working -- no applicable method for group_map applied to object of class "multidplyr_party_df"

[JasonAizkalns]

Consider this example:

library(dplyr)
library(multidplyr)

cluster <- new_cluster(2)

# Works fine...
mtcars %>%
  group_by(cyl) %>%
  group_walk(~ write.csv(.x, paste0(.y, ".csv"))

# This fails...
mtcars %>%
  group_by(cyl) %>%
  partition(cluster) %>%
  group_walk(~ write.csv(.x, paste0(.y, ".csv"))

# Error in UseMethod("group_map") : 
#  no applicable method for 'group_map' applied to an object of class "multidplyr_party_df"

Maybe because group_walk is experimental?

[TimBMK]

I have a similar issue with slice_max(), which is not experimental. It seems not all functions are supported (yet) by multidplyr?

Edit: ... and apparently the same goes for distinct()?

[hadley]

Thanks for the suggestion! Will definitely consider it when I'm next working on multidplyr.