recipe_julia.yml produces unusable plot

[axel-lauer]

recipe_julia.yml produces a plot with an unusable color scale (0 ... 1e20 K) for average surface temperature. It seems missing values (1e20) are not recognized as such.

Output looks like this:

[valeriupredoi]

sounds to me that it's not doing proper masking, what you reckon @jhardenberg ? 🍺

[remi-kazeroni]

I'm adding recipe_julia.yml to the list of "broken recipes". The log file ends with "Recipe run was successful" but since a few releases, we still get the same unusable plot. See here for v2.9.0. Since the plot is identical for different versions of ESMValTool, the comparison tool does not issue a warning.

julia is probably not the easiest dependency to handle for us. I think it would be nice that the only plot generated with a Julia diagnostic would give meaningful results. By drawing attention to this recipe, I hope someone would find the time to take a look at this issue.

[valeriupredoi]

found a fix for this, opening a draft PR

[bouweandela]

It looks like the Julia software stack is finally starting to become useful for climate data: https://github.com/mauro3/Julia-intro-for-geoscience-EGU2024/blob/main/notebooks/geo-ecosystem.ipynb. Maybe we could update the example recipe so it uses these libraries and that will make it more attractive.