row_number() fails in the some corner cases

[vpanfilov]

row_number() windowed function fails in the corner case with 0x0 tibble:

library(tidyverse)

tibble() %>% mutate(a = row_number())
#> Error in mutate_impl(.data, dots) :
#>   Evaluation error: argument "x" is missing, with no default.

UPD: Found another misleading behaviour:

library(tidyverse)

tibble(a = 1) %>%
  mutate(b = row_number())
#> # A tibble: 1 x 2
#>       a     b
#>   <dbl> <int>
#> 1    1.     1

tibble(a = 1) %>%
  mutate(b = dplyr::row_number())
#> Error in mutate_impl(.data, dots) : 
#>   Evaluation error: argument "x" is missing, with no default.

[batpigandme]

What about using:

suppressPackageStartupMessages(library(dplyr))
tibble::tibble() %>% tibble::rowid_to_column("a")
#> # A tibble: 0 x 1
#> # ... with 1 variable: a <int>

Created on 2018-03-21 by the reprex package (v0.2.0).

[vpanfilov]

@batpigandme, thanks! Will use rowid_to_column in my code.

Added another misleading behaviour of the row_number() to the first post.

[krlmlr]

Thanks. The second example does work in the dev version of dplyr:

library(tidyverse)

tibble() %>%
  mutate(b = dplyr::row_number())
#> Error in mutate_impl(.data, dots): Evaluation error: argument "x" is missing, with no default.

tibble(a = integer()) %>%
  mutate(b = dplyr::row_number())
#> # A tibble: 0 x 2
#> # ... with 2 variables: a <int>, b <int>

tibble(a = 1) %>%
  mutate(b = dplyr::row_number())
#> # A tibble: 1 x 2
#>       a     b
#>   <dbl> <int>
#> 1    1.     1

Created on 2018-03-23 by the reprex package (v0.2.0).

[krlmlr]

@romainfrancois: Can you please take a look? The first example in my last post with the empty tibble should still work.

[lock]

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/