Inconsistent results from names() and dimnames() on database backed tbl()

tidyverse
1

As mentioned here you can get the column names from a dbplyr tbl using colnames() but not using names(). ... that's cool. But, does anyone know offhand why/how is it this way? colnames() uses the primitive dimnames() and names() itself is a primitive. It doesn't look like S3 dispatch is happening for either.

2

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

3

Hello @RussellPierce! Apologies for the delayed response here. This is a great question!!

The reason for this is that dimnames uses S3 dispatch! You can see this with the following:

dimnames.data.frame
#> function (x) 
#> list(row.names(x), names(x))
#> <bytecode: 0x7f814ab3cfc0>
#> <environment: namespace:base>
dbplyr:::dimnames.tbl_lazy
#> function (x) 
#> {
#>     list(NULL, op_vars(x$ops))
#> }
#> <environment: namespace:dbplyr>

Created on 2020-03-20 by the reprex package (v0.3.0)

I don't know a whole lot of the specifics about how this works in base R, but perhaps others will! Unfortunately this thread is closed already :man_facepalming: So perhaps a new discussion / thread on how base R primitives do S3 dispatch / how to discern that would be worthwhile? :man_shrugging: