This is basically a resubmission of tidyr 0.5.0 - I've fixed a small buglet that caused a failure on R-oldrel. Sorry for only noticing it after the release! (I've started routinely testing on oldrel so hopefully this will be less of a problem in the future)
- local OS X install, R 3.2.3
- ubuntu 12.04 (on travis-ci), R 3.2.3
- win-builder (devel and release)
0 errors | 0 warnings | 2 note
-
License components with restrictions and base license permitting such: MIT + file LICENSE
-
Missing or unexported object: 'dplyr::everything'
This is a bit of a hack to get tidyr working with both dplyr 0.4 and dplyr 0.5. It'll disappear once dplyr 0.5 is out.
I ran R CMD check
on all 73 reverse dependencies.
Results at https://github.com/hadley/tidyr/tree/master/revdep/.
Problems are summarised below. The are 8 problems caused by a deliberate change to tidyr. I notified authors on May 26 and again today.
A number of packages fail because tidyr now returns objects of class c("tbl_df", "tbl", "data.frame"), and the $.tbl_df method throws error instead of returning NULL for non-existant variables:
- broom: checking examples ... ERROR
- ezsummary: checking tests ... ERROR
- ggmcmc: checking examples ... ERROR
- heemod: checking tests ... ERROR
- pixiedust: checking re-building of vignette outputs ... WARNING
- quickpsy: checking examples ... ERROR
- R6Frame: checking tests ... ERROR
- RNeXML: checking examples ... ERROR
Other problems:
-
pmc: checking tests ... ERROR I'm not sure what the cause of this one is
-
rgho: checking examples ... ERROR Caused by assuming that column names are characters (which they should be!)
-
ggRandomForests: checking examples ... ERROR This same error is present on CRAN
-
glycanr: checking examples ... ERROR Examples is missing check for installation of suggested package
-
tigger: checking re-building of vignette outputs ... WARNING Vigentte is missing check missing check for installation of suggested package