7

I'm building an R package and have run into a perplexing warning during R CMD check:

* checking for missing documentation entries ... WARNING
Undocumented data sets:
  ‘.Random.seed’

The package has one small data set, which is documented. I'm using R 3.1.1 and RStudio 0.98.1062 on OS X Yosemite, but I get the same error on Windows 7 (and from CRAN). The project also has a vignette that is built with knitr. devtools etc. are all up to date. The file '.Random.seed' doesn't exist in the "data" folder before building, and my reasoning is that it's getting transiently written to disk during the build process by...something. I tried adding '.Random.seed' to .Rbuildignore without success, presumably because it doesn't exist when the build process begins.

Has anyone encountered this before?

Phil
  • 7,287
  • 3
  • 36
  • 66
schmolze
  • 71
  • 3
  • 2
    Do you have `.Random.seed` in your code anywhere? Please take the time to create a [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) to make it easier to help you. – MrFlick Nov 12 '14 at 16:02
  • put it on github or something – rawr Nov 12 '14 at 16:16
  • No references to .Random.seed in my code anywhere. It's at https://github.com/schmolze/ega if anyone wants to try building it. – schmolze Nov 12 '14 at 16:57

2 Answers2

14

Ran into this problem as well. You have almost certainly solved it by now, but I'll post an answer just in case somebody else hits the same problem. At some point, you generated a random number or set the seed in the creation of the Rdata file (or at least that's what happened to me). Simply load the workspace from you data folder, and rm(.Random.seed). Save it. You're done. Easy as pie.

http://www.inside-r.org/r-doc/base/set.seed

Tad Dallas
  • 1,179
  • 5
  • 13
0

I had a similar problem when I was uploading the my.csv dataset, if anyone faces a similar problem the answer is inhere.

Seyma Kalay
  • 2,037
  • 10
  • 22