R devel - Dec 2019 - how to check as CRAN with alternative BLAS?

One of my packages is slated to be archived from CRAN due to failures when
the ATLAS BLAS is used. I am unable to replicate the error on my machine
under R 3.6.1 using the atlas library from ubuntu (seems to be 3.10.2-9,
while the good professor is using 3.10.3 per
https://www.stats.ox.ac.uk/pub/bdr/Rblas/README.txt ). I also tried the
rocker/r-base with R 3.6.2
and /usr/lib/x86_64-linux-gnu/atlas/libblas.so.3.10.3 (Dockerfile here:
https://gist.github.com/shabbychef/efaa048f0f715dcf89fd39f2e28a402d ), but
was unable to replicate the error there. Is there some way to *really* test
as CRAN-with-atlas? I would prefer a dockerized solution, or failing that,
a travis CI recipe.

thanks,

-- 

--sep

	[[alternative HTML version deleted]]

On 29 December 2019 at 16:39, steven pav wrote:
| One of my packages is slated to be archived from CRAN due to failures when
| the ATLAS BLAS is used. I am unable to replicate the error on my machine
| under R 3.6.1 using the atlas library from ubuntu (seems to be 3.10.2-9,
| while the good professor is using 3.10.3 per
| https://www.stats.ox.ac.uk/pub/bdr/Rblas/README.txt ). I also tried the
| rocker/r-base with R 3.6.2
| and /usr/lib/x86_64-linux-gnu/atlas/libblas.so.3.10.3 (Dockerfile here:
| https://gist.github.com/shabbychef/efaa048f0f715dcf89fd39f2e28a402d ), but
| was unable to replicate the error there. Is there some way to *really* test
| as CRAN-with-atlas? I would prefer a dockerized solution, or failing that,
| a travis CI recipe.

That CRAN checks cannot always be replicated outside of CRAN is a real issue.

This has been brought up before, and some of us wanted to see about improving
it, possibly relying on Docker---but as you know life is short, spare time is
generally missing, and other shiny things can distract us.  So nothing to
report yet. Which is a bummer.

Dirk

-- 
http://dirk.eddelbuettel.com | @eddelbuettel | edd at debian.org

Give a try to "rhub":

https://cran.r-project.org/web/packages/rhub/index.html

https://cran.r-project.org/web/packages/rhub/vignettes/local-debugging.html

Which allows to try different os, and system configurations :)

Hope it works!

El lunes, 30 de diciembre de 2019, Dirk Eddelbuettel <edd at debian.org>
escribi?:
>
> On 29 December 2019 at 16:39, steven pav wrote:
> | One of my packages is slated to be archived from CRAN due to failures
> when
> | the ATLAS BLAS is used. I am unable to replicate the error on my machine
> | under R 3.6.1 using the atlas library from ubuntu (seems to be 3.10.2-9,
> | while the good professor is using 3.10.3 per
> | https://www.stats.ox.ac.uk/pub/bdr/Rblas/README.txt ). I also tried the
> | rocker/r-base with R 3.6.2
> | and /usr/lib/x86_64-linux-gnu/atlas/libblas.so.3.10.3 (Dockerfile here:
> | https://gist.github.com/shabbychef/efaa048f0f715dcf89fd39f2e28a402d ),
> but
> | was unable to replicate the error there. Is there some way to *really*
> test
> | as CRAN-with-atlas? I would prefer a dockerized solution, or failing
> that,
> | a travis CI recipe.
>
> That CRAN checks cannot always be replicated outside of CRAN is a real
> issue.
>
> This has been brought up before, and some of us wanted to see about
> improving
> it, possibly relying on Docker---but as you know life is short, spare time
> is
> generally missing, and other shiny things can distract us.  So nothing to
> report yet. Which is a bummer.
>
> Dirk
>
> --
> http://dirk.eddelbuettel.com | @eddelbuettel | edd at debian.org
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
	[[alternative HTML version deleted]]