tibble() and zero-row data frames

[krlmlr]

tibble() and zero-row data frames

[imanuelcostigan]

A regression in v1.1?

> tibble(~a, ~b)
# Error: Each variable must be a 1d atomic vector or list.
# Problem variables: '~a', '~b'

Is this desired behaviour? I am trying to understand the following:

> tibble(a = numeric(), b = character())
# A tibble: 0 x 2
# ... with 2 variables: a <dbl>, b <chr>

But

> tibble(a = NULL, b = NULL)
# Error: Each variable must be a 1d atomic vector or list.
# Problem variables: 'a', 'b'

Related to: imanuelcostigan/RSQLServer#73

[imanuelcostigan]

To be clear, running these two equivalent calls (in SQL space) result in two different outcomes:

DBI: dbGetQuery(ss$con, "SELECT * FROM AIRLINES WHERE carrier = 'ZZ'")

# [1] carrier name   
# <0 rows> (or 0-length row.names)

dplyr: tbl(ss, "airlines") %>% filter(carrier == 'ZZ')

# Error: Each variable must be a 1d atomic vector or list.
# Problem variables: 'carrier', 'name'

I'm trying to work out what I should expect from upstream dependencies.

[imanuelcostigan]

Pinging @krlmlr. Should I open a new ticket?

[krlmlr]

I think this is intended, but please open a new ticket. What type should the new NULL columns have? Also, tibble() is the same as data_frame() now, but was an alias for frame_data() in 1.0.

[github-actions]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.