dplyr 1.0.8

[romainfrancois]

We're about to release dplyr 1.0.8 and we've identified that the current huxtable fails against it with:

── After ─────────────────────────────────────────────────────────────────────────────────────────────────────
> checking tests ... ERROR
  See below...

> checking dependencies in R code ... NOTE
  Namespaces in Imports field not imported from:
    ‘R6’ ‘xml2’
    All declared Imports should be used.

── Test failures ─────────────────────────────────────────────────────────────────────────────── testthat ────

> library(testthat)
> library(huxtable)
> 
> 
> we_are_in_R_CMD_check <- TRUE
> test_check("huxtable")
# weights:  9 (4 variable)
initial  value 35.155593 
iter  10 value 23.131901
final  value 23.129234 
converged
══ Skipped tests ═══════════════════════════════════════════════════════════════
• 0-length tables don't work in flextable yet (1)
• Awaiting ftExtra improvements (1)
• Couldn't unload dplyr namespace (2)
• Not testing, code doesn't play well with R CMD check (3)
• On CRAN (16)

══ Failed tests ════════════════════════════════════════════════════════════════
── Failure (test-dplyr.R:56:3): mutate, mutate_ and transmute work ─────────────
bold(ht4)[, 1] not equivalent to c(FALSE, FALSE, FALSE).
1 element mismatch

[ FAIL 1 | WARN 2 | SKIP 23 | PASS 1207 ]
Error: Test failures
Execution halted

1 error x | 0 warnings ✓ | 1 note x

I've tried to test with the dev version of huxtable, but I got more errors (most non related).

[hughjonesd]

Okay. This is a test that is being a little too finicky about the column order which it expects in transmute(). I think I'll just remove the test. But, I note that it's driven by transmute() doing columns in a new order, and debate on this is ongoing in the linked issue. I don't need to do a new release if you decide not to change the column order, so I'll watch https://github.com/tidyverse/dplyr/issues/6086 and see what happens.

[hughjonesd]

Closing. Reopen if problems recur.