Yura said to look at code review

[vjugor1]

Structure

• Please, make the project structure as in crops or wind
• Add Dockerfile, avoid package installation from conda - use requirements.txt (Examples 2 from here -- generate requiremets.txt with pip freeze)
• Fix README.md to be displayed in the repository
• Add results folder that would contain figure that you are saving and tables (.csv) - push them there also

Notebooks

• Please, cast best_models_nested to a pandas.DataFrame and save it to .csv
• Could you please fix the warnings: FutureWarning: Index.ravel returning ndarray is deprecated; in a future version this will return a view on self. and UserWarning: CRS not set for some of the concatenation inputs. Setting output's CRS as WGS 84 (the single non-null crs provided).

correlation.py

• Place "2015_2022" to the arguments here and further
• Place num_models into the arguments of this function
• Add progress bars in the loops of the optimal_ensemble function

process.py

• Are you sure that this coordinate transform is conducted correctly? Please, give here or in youtrack ranges of original lon axes and transformed ones. I am suspicious, since we were transforming coordinates in a different way (for cmips)
• Shouldn't it work a more elegant way like era.sel(time=slice(year, year+1))?
• Should it also return era_year_clip? Otherwise, it look inconsistent with the docstring of this function
• From here to there can be placed into a separate utility function - would be more nicer
• The same elegance suggestion here
• Shouldn't this function return something as well? Or update docstring (I guess better to update)
• Same here and here and here and here

Data

• Could you please add to this repository the code that you were using for CMIP downloading? Together with context.jsons, if you saved them
• Could you also provide in the README.md locations of the data that you were using

Thanks! That's a good code

[DariaTan]

Add progress bars in the loops of the optimal_ensemble function

I didn't add it, the cell executes a sec or less

Are you sure that this coordinate transform is conducted correctly? Please, give here or in youtrack ranges of original lon axes and transformed ones. I am suspicious, since we were transforming coordinates in a different way (for cmips)

Checked, works fine
original lon axes: -180.0 -179.8 -179.5 -179.2 ... 179.2 179.5 179.8

transformed ones: 0.0 0.25 0.5 0.75 1.0 ... 359.0 359.2 359.5 359.8

Shouldn't it work a more elegant way like era.sel(time=slice(year, year+1))

You are right, it looks much better
However, this slice takes both years (year and year+1) in not very pythonic way)
I did it with era.sel(time=year) for the single year of course

[vjugor1]

Nice, thanks!