Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

pinning the R version to an MRAN date #245

Closed
fomightez opened this issue Feb 22, 2018 · 9 comments
Closed

pinning the R version to an MRAN date #245

fomightez opened this issue Feb 22, 2018 · 9 comments
Labels
enhancement needs: discussion reproducibility enabling scientific reproducibility with r2d

Comments

@fomightez
Copy link
Contributor

Could the version of R installed gets tied to the MRAN snapshot with the data specified in runtime.txt as well? Right now with the method for builds without using a Dockerfile, the version of R installed is pinned to one date, see https://github.com/jupyter/repo2docker/blob/master/repo2docker/buildpacks/r.py#L132 .

In contrast, Bioconductor releases are tied to certain versions of R. This can be seen from past and present release notes:

Note from an older release

"The current release of Bioconductor is version 3.4; it works with R version 3.3.1. Users of older R and Bioconductor users must update their installation to take advantage of new features and to access packages that have been added to Bioconductor since the last release."

(Source of similar information still accessible here)

Current release note:

"The current release of Bioconductor is version 3.6; it works with R version 3.4.3. Users of older R and Bioconductor must update their installation to take advantage of new features and to access packages that have been added to Bioconductor since the last release."

And so if somone wants to run the packages associated with a particular past Bioconductor release as they were, they'd need the specific version of R.

I don't know if it would be best to also pin the RStudio version as well or not?

@willingc willingc added user-support Questions on using repo2docker needs: discussion labels Mar 13, 2018
@willingc
Copy link
Contributor

Thanks for raising the questions here. Any other R folks that have thoughts on this?

@yuvipanda
Copy link
Collaborator

yuvipanda commented Mar 13, 2018 via email

@jzf2101
Copy link
Contributor

jzf2101 commented Jun 28, 2018

I think this is already a feature- can we close? @choldgraf ?

@yuvipanda
Copy link
Collaborator

This isn't implemented yet, unfortunately. Setting the MRAN date only sets the version of libraries being installed, not R itself.

@jzf2101
Copy link
Contributor

jzf2101 commented Jun 28, 2018

@betatim
Copy link
Member

betatim commented Jun 30, 2018

Using runtime.txt does not influence what version of R will be installed, only what version of the R packages.

@jzf2101 jzf2101 added the reproducibility enabling scientific reproducibility with r2d label Jul 2, 2018
@HeidiSeibold
Copy link

HeidiSeibold commented Sep 10, 2018

Maybe it helps to take a look at what the Rocker folks are doing?

https://github.com/rocker-org/rocker-versioned

As an example, this gets you R version 3.3.0:

https://github.com/rocker-org/rocker-versioned/blob/master/r-ver/3.3.0/Dockerfile
(Look for R_VERSION=${R_VERSION:-3.3.0})

@nuest
Copy link
Contributor

nuest commented Sep 17, 2019

#772 should close this (going back as far as R 3.4).

Also, the rocker/binder image has older versions if a user adds a Dockerfile to the repo.

@nuest
Copy link
Contributor

nuest commented Nov 5, 2019

@betatim Just came across this.. can be closed, right?

@betatim betatim closed this as completed Nov 8, 2019
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
enhancement needs: discussion reproducibility enabling scientific reproducibility with r2d
Projects
None yet
Development

No branches or pull requests

7 participants