search

NWTlter / NWT_CLM

Workflow scripts for running point simulations of CLM at Niwot Ridge using Tvan forcing data
1 stars 7 forks source link

adding r packages - except reddyproc - to conda environment #32

Closed hhollandmoritz closed 2 years ago

hhollandmoritz commented 2 years ago

@wwieder I've updated the conda environment to include R packages. You can see if this solves the certificate problem by testing this branch.

The conda environment will probably take a little while to install (it often does with R packages from conda forge because there are so many).

I haven't looked into installing reddyproc via conda yet, so for right now, it will just install itself at the top of the script (per the script installation code at the top of the script).

To make sure that Rstudio sees the correct environment, you can

  1. activate the conda environment,
  2. start rstudio from the command line - you should be able to type rstudio in the terminal and have it start,
  3. Once Rstudio is up and running, check to make sure that R.home() and .libPaths() are referencing the conda version of R (see this post for more details).
wwieder commented 2 years ago

thanks for this suggestion @hhollandmoritz . The previous conda environment was created for the jupyter notebook scripts that are in the notebooks directory and used for analyzing he CLM output. Is it necessary to add all these r libraries to the ctsm-py environment? Are we any better off creating a second .yml file that includes these r-specific libraries for the preprocessing of tvan data in steps 1-2 of this workflow, or is it better practice to have all this together?

hhollandmoritz commented 2 years ago

My personal thoughts are that it's better to have it all in one place as it's more repeatable and then it comes as one package. But @scelmendorf may have different opinions and I'd say she's the expert on reproducible science. I think typically, folks use only one environment per project, but I'm not opposed to having multiple ones as long as it's clear in the readme what each is for. It's certainly not hard to create two sets of yaml directives.

More importantly, if this doesn't actually fix your "no certificate found" problem, then I'd say it's probably not worth implementing since that was the whole impetus for doing it in the first place.

hhollandmoritz commented 2 years ago

Did the conda environment fix it? @wwieder

wwieder commented 2 years ago

qualified yes. I also had issues with the ameriflux data download, but I think it's for a different issue. Now I can seem to log onto cheyenne to test out the new input data...

On Tue, Nov 9, 2021 at 11:56 AM Hannah Holland-Moritz < @.***> wrote:

Did the conda environment fix it? @wwieder https://github.com/wwieder

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/NWTlter/NWT_CLM/pull/32#issuecomment-964441666, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB5IWJHXOQBHFBJBIHMHJADULFVGJANCNFSM5HVKPFZQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.