[R-pkg-devel] Failing incoming tests on 'stringsAsFactors = TRUE'

Trent McDonald tmcdon@|d @end|ng |rom we@t-|nc@com
Tue Mar 31 00:42:15 CEST 2020


Trying to update a package on CRAN.  It passes all the following tests:
Local (r CMD --as-cran, R 3.6.2), devtools::check_win_dev,
devtools::check_rhub, and my personal .yml file on travis-ci.org   But, the
incoming CRAN tests last week for windows fail (linux test throw different
warning that I can fix).  This is the log file:

* using log directory 'd:/RCompile/CRANincoming/R-devel/SDraw.Rcheck'
* using R version 4.0.0 alpha (2020-03-26 r78078)
* using platform: x86_64-w64-mingw32 (64-bit)
* using session charset: ISO8859-1
* checking for file 'SDraw/DESCRIPTION' ... OK
* checking extension type ... Package
* this is package 'SDraw' version '2.1.10'
* checking CRAN incoming feasibility ... Note_to_CRAN_maintainers
Maintainer: 'Trent McDonald <tmcdonald using west-inc.com>'
* checking package namespace information ... OK

--- snip ---- multiple checks, all OK

* checking 'build' directory ... OK
* checking DESCRIPTION meta-information ... WARNING
Warning message:
In options(stringsAsFactors = TRUE) :
  'options(stringsAsFactors = TRUE)' is deprecated and will be disabled
 ERROR
Warning message:
In options(stringsAsFactors = TRUE) :
  'options(stringsAsFactors = TRUE)' is deprecated and will be disabled
* DONE
Status: 1 ERROR, 1 WARNING

Course forward unknown.  Googling unproductive.  I searched the package
source for the use of 'stringsAsFactors'. I never set the option nor call
data.frame with it nor read.csv nor use stringsAsFactors anywhere.  In my
local .Rprofile, I set 'options(stringsAsFactors=FALSE)', but that should
not matter, correct?  Suggestions?

My local version is 3.6.2.  Source code link :
https://github.com/tmcd82070/SDraw


*Trent McDonald*
* Senior Statistician*

	[[alternative HTML version deleted]]



More information about the R-package-devel mailing list