Diego Hernangómez Herrero identified my problem:
My .Rbuildignore file in the package root directory contained
'MaddisonCountries.rda'. I thought that entry in .Rbuildignore would
apply only to the a copy in the package root directory and would not
affect the copy in data. Clearly I was wrong.
The code in my vignette only generates that file when there's an
appropriately named *.xlsx file in inst/exdata. Both the *.xlsx and
*.rda files are in .gitignore and so are not on GitHub. That trick means
that the vignette does not write anything on standard checks on GitHub,
CRAN, etc.
I can accomplish the same thing without this problem and still have the
vignette work as intended by writing the new data file with a different
name. The file would then have to be renamed before replacing the old
file in data/ with the new.
Thanks again to Diego Hernangómez Herrero and thanks to all others who
spent time looking at this and thinking about it.
Sincerely,
Spencer Graves
On 9/24/25 19:01, Spencer Graves wrote:
Hi, John:
On 9/24/25 16:27, John Benninghoff wrote:
Hi Spencer, I tried downloading your GitHub repo. I wasn’t able to
replicate your R CMD check results. I did find issues with the
DESCRIPTION file, which I fixed by removing the blank “Imports” and
running desc::desc_normalize(). After that, running
rcmdcheck::rcmdcheck() returned 0 errors, warnings, and notes.
Thanks.
I got what you got with "rcmdcheck::rcmdcheck()".
Sadly, GitHub Actions still reported the previous warning after I
committed the change you suggested to DESCRIPTION AND locally using the
process I found recommended in "R Projects":
library(devtools)
document() # add
load_all()
check()
I'm currently trying to develop a toy package using roxygen2:
https://github.com/sbgraves237/roxygenData1
This toy package currently consists only of boilerplate with zero content.
Thanks for your help. roxygen2 sounds great, but it cannot be used if a
problem like this cannot be fixed with a finite amount of effort ;-)
Spencer
On Sep 24, 2025, at 3:32 PM, Spencer Graves
<spencer.gra...@effectivedefense.org> wrote:
Hi, Michael et al.:
On 9/24/25 11:59, Michael Chirico wrote:
{ggplot2} has a number of data sets, you might read over their sources:
https://github.com/tidyverse/ggplot2/blob/main/R/data.R <https://
github.com/tidyverse/ggplot2/blob/main/R/data.R>
Thanks. I don't see anything there that is different from the example
in their book, which is extracted from the similar data.R file in
their tidyr package, which I reviewed earlier.[2]
I've deleted virtually everything else in that package and still get
that same error.
Suggestions?
Spencer Graves
p.s. I've done several web searches on this issue, found discussions
on StackOverflow that may have worked a few years ago, but roxygen2
seems to have changes since then. I posted a similar question to
StackOverflow, that generated a few comments but nothing that seemed
to fix this problem. I reported an "issue" to the R Packages book on
GitHub last week, so far with nothing:
https://github.com/hadley/r-pkgs/issues/1106
I tried to use roxygen2 with a different project a year ago and gave
up after being unable to figure out how to use it. >
On Wed, Sep 24, 2025 at 5:48 AM Spencer Graves
<spencer.gra...@effectivedefense.org
<mailto:spencer.gra...@effectivedefense.org>> wrote:
Hello, All:
How does one document a dataset with roxygen2?
I've recently read Wickham and Bryan, R Packages (2e).[1] I've
tried to
do what they say there, and R CMD check is reporting:
Variables with usage in Rd file 'MaddisonCountries.Rd' but
not in
code:
‘MaddisonCountries’
This is in "https://github.com/sbgraves237/SS4Maddison <https://
github.com/sbgraves237/SS4Maddison>".
Suggestions? Thanks, Spencer Graves
p.s. Wickham and Bryan Section 7.1.2 says, "Objects in data/ are
always
effectively exported .... . This means that they must be
documented. ...
[Y]ou document the name of the dataset and save it in R/. For
example,
the roxygen2 block used to document the who data in tidyr is
saved in
R/data.R ... . There are two roxygen tags that are especially
important
for documenting datasets: @format ... @source
They give an example of a "who" dataset in tidyr. I've studied that
example without finding what I'm doing wrong.[2]
In addition, I need to specify Roygen and "LazyData: true" in
DESCRIPTION.
[1] Wickham and Bryan, R Packages online at:
https://r-pkgs.org/ <https://r-pkgs.org/>
Chapter 7 deals with Data:
https://r-pkgs.org/data.html <https://r-pkgs.org/data.html>
[2] "data.R" in tidyr is available at:
https://github.com/tidyverse/tidyr/blob/main/R/data.R <https://
github.com/tidyverse/tidyr/blob/main/R/data.R>
______________________________________________
R-package-devel@r-project.org <mailto:R-package-devel@r-
project.org>
mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel <https://
stat.ethz.ch/mailman/listinfo/r-package-devel>
______________________________________________
R-package-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
______________________________________________
R-package-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
______________________________________________
R-package-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-package-devel
