[R-pkg-devel] tibbles are not data frames

[Hadley Wickham]

On Tue, Sep 26, 2017 at 12:15 PM, Patrick Perry <pperry at stern.nyu.edu> wrote:
> Pro ignoring x[,1,drop=TRUE]:
> (1) it forces users to write consistent code for extracting a vector from a
> data frame
>
> Con:
> (1) functions that accept both matrices and data frames might break
> (x[[j]][i] doesn't work for a matrix)

I generally think that it's better to keep matrices and data frame
completely separate, but point taken.

> (2) functions that use the access pattern x[i,j,drop = TRUE] will break

This seems pretty rare, and I don't think anyone has complained about it yet.

I don't love adding support for drop = TRUE because it makes [.tibble
type-unstable, but maybe it's reasonable to do so in order to slightly
improve backward compatibility. I've filed an issue so we consider it
for the next major release:
https://github.com/tidyverse/tibble/issues/311

> Perhaps a bigger issue with tibbles is that they don't let you index with
> row names:
>
>> y <- tibble(x = letters)
>> rownames(y)
>  [1] "1"  "2"  "3"  "4"  "5"  "6"  "7"  "8"  "9"  "10" "11" "12" "13" "14"
> "15"
> [16] "16" "17" "18" "19" "20" "21" "22" "23" "24" "25" "26"
>> y[rownames(y)[c(1,5,9,15,21)],]
> # A tibble: 5 x 1
>       x
>   <chr>
> 1  <NA>
> 2  <NA>
> 3  <NA>
> 4  <NA>
> 5  <NA>

I'd argue that this is not as big as an issue, as I have no
recollection of anyone complaining about it.

Hadley

-- 
http://hadley.nz