search

ESMValGroup / ESMValTool

ESMValTool: A community diagnostic and performance metrics tool for routine evaluation of Earth system models in CMIP
https://www.esmvaltool.org
Apache License 2.0
217 stars 127 forks source link

[R dependencies] Relax restrictive pin on `r-akima` #3563

Closed valeriupredoi closed 6 months ago

valeriupredoi commented 6 months ago

In https://github.com/ESMValGroup/ESMValTool/issues/2695 a hard upper bound was set on r-akima - that is constraining us from upgrading support to Pyton 3.12, see https://github.com/ESMValGroup/ESMValTool/pull/3501 - is it safe to remove that pin, or, at least, do a not versions pin? @ESMValGroup/technical-lead-development-team akimbo-matata :lion: :grin:

sloosvel commented 6 months ago

Removing the pin will break the recipes, but they are unmaintained anyway

valeriupredoi commented 6 months ago

that's a truly fatalistic approach @sloosvel :laughing: Can you pls try running one of the recipes with https://github.com/ESMValGroup/ESMValTool/pull/3564 see if its still breaks it? :beer:

sloosvel commented 6 months ago

Well all three recipes fail, but I am not sure if it's related to that package in particular:

[1] "saving NetCDF climatologies..."
Error in (models_dataset == reference_dataset) && (models_exp == reference_exp) :
  'length = 2' in coercion to 'logical(1)'
Calls: which
Execution halted

It seems like a different error than in the past. This is the version that is being used without the pin:

r-akima                   0.6_3.4           r43h61816a4_2    conda-forge
valeriupredoi commented 6 months ago

@sloosvel you very kind, thank you for testing those wRetched recipes! That is indeed the latest version https://anaconda.org/conda-forge/r-akima - and if you are happy with that error not coming out of r-akima's arse then we can merge this :beer:

valeriupredoi commented 6 months ago

actually I think that error is what we've seen from those recipes for a while now, and had nobody to fix it