search

CredibilityLab / groundhog

Reproducible R Scripts Via Date Controlled Installing & Loading of CRAN & Git Packages
https://groundhogr.com/
GNU General Public License v3.0
78 stars 4 forks source link

groundhog.library failing to read connection #78

Closed alexfradera closed 2 years ago

alexfradera commented 2 years ago

Today I have been testing out knitting some .rmd code with calls to libraries within it, and for the first time trying to update these to use groundhog.library instead. This have been proving quite difficult and I might separately document other parts of the issue (eg constantly looping "To ensure you are actively answering, only 'uninstall' and 'keep' are accepted as responses") and in the course of which I manually uninstalled and reinstalled stringi, which seems to be something that the knit process itself relies on.

I don't know if that triggered something, but groundhog.library stopped running in any context, in any project, just as code - it just read

> groundhog.library("tidyverse", "2021-09-30")
Error in readRDS(toc.path) : error reading from connection

When I changed the groundhog path away from its normal one to another one, it began working again. Then it continued to work when I moved it back. However, I needed to repeat the process for different projects. And when it comes to the rmd knit, it seems impervious:

quitting from lines 13-58 (./SR_Report_Body.rmd)
Error in readRDS(toc.path) : error reading from connection
Calls: <Anonymous> ... groundhog.library.single -> get version -> toc -> load.cran.toc -> readRDS

I'm trying to figure out if I can find a way around this, or whether there are issues with trying to call groundhog from within a .rmd?

urisohn commented 2 years ago

Working on rmd files, with R Studio, while using groundhog is indeed tricky. The reason, I think, is that R Studio loads packages on its own, even pkgs not references in your scripts, making version control hard. I think the new version of groundhog, not yet released, will be more resistant to this R Studio behavior. Version v.2.0.0 also will work with github and gitlab packages and am currently workign on addressing conflicts that arise in that context. Will test .rmd file conflicts next. I think/hope that the new release will be ready in a matter of weeks, so maybe I would recommend to not use groundhog in your rmd files until that happens.

alexfradera commented 2 years ago

That's great, thank you, I'll take a different tack.

urisohn commented 2 years ago

Now that v2.0.0 is out, this should work for you. If you try it out, let me know how it goes.

alexfradera commented 2 years ago

Hi Uri,

Great thanks for the update!

My thesis is going to the printer in less than a week so for my sanity I will stick to things as-is, but will take a look at it afterward. Thanks for looking into it, getting that working makes it massively helpful!

Best Alex

On Wed, 20 Jul 2022 at 09:06, Uri Simonsohn @.***> wrote:

Now that v2.0.0 is out, this should work for you. If you try it out, let me know how it goes.

— Reply to this email directly, view it on GitHub https://github.com/CredibilityLab/groundhog/issues/78#issuecomment-1189960777, or unsubscribe https://github.com/notifications/unsubscribe-auth/ASKXFBOVEQ2KAROAO3NVCSLVU6XQ5ANCNFSM5WRZXZ2Q . You are receiving this because you authored the thread.Message ID: @.***>

urisohn commented 2 years ago

ok, i understand. Good luck!