Is there an equivalent of plyr's .inform in dplyr?

Question

plyr has a very convenient option, .inform = TRUE , which lets user know which pieces in the split-apply-combine process fails.

Is there an equivalent way to know which piece fails with dplyr 's group_by ?

A minimal example: group 1 in the following df is the group causing problem, but dplyr error message doesn't reveal that information.

df <- data.frame(group = c(1, 1, 2, 2, 2))
df %>% group_by(group) %>% mutate(value = c(10, 10, 10))
# Error: incompatible size (3), expecting 2 (the group size) or 1

Answer 1

I don't know a way within dplyr , so this may be a bit orthogonal to what you're looking for, but here's a potential approach using split and purrr::possibly .

library(purrr); library(dplyr)

set_3_tens <- function(df) { df %>% mutate(value = c(10,10,10)) }

df %>%
  split(.$group) %>%
  map(
    possibly(
      set_3_tens, otherwise = "Problem in group"))

$`1`
[1] "Problem in group"

$`2`
  group value
1     2    10
2     2    10
3     2    10