Add dummy naming scheme for AWI model about xmip HOT 4 CLOSED

First, I just want to say that this is an amazing tool. Glad to see how nicely this tool integrates with intake-esm. Thank you for putting it together! I am planning on testing it on Cheyenne in the coming days.

From a user perspective, I was wondering if it is worth moving the dim_name_dict into a YAML or JSON config file (persisted outside the package (~/config/cmip6_preprocessing.yaml) maybe? or some other location) that gets read by the package?
https://github.com/jbusecke/cmip6_preprocessing/blob/bcc9c108ecae72a8460f012b37756af691cd2985/cmip6_preprocessing/preprocessing.py#L196-L200

Having the content of the dictionary in a config file would probably make it easy for the user to extend the dictionary for their own use cases without needing to make changes in the code base itself. Thoughts?

from xmip.

Thanks for the kind words @andersy005. That is a great suggestion.

I agree that a yaml would be easier to handle, but my initial hope was that we could build a comprehensive database of all the 'quirks` in CMIP6 over time, so eventually the user does not have to configure this anymore (thus I would store it in the package.

What would be the use-case for a user-customized file in your opinion?

from xmip.

Just picking back up on this. How about we have a 'built in' database (cmip6), but then we could check for another file in the user directory and overwrite the existing one?
I am not that familiar with handling yaml files, but there must be an easy way to parse to a dict, right?

from xmip.

I have recently changed the way we use the internal renaming to a universal renaming dictionary. This is easier to manage, but perhaps it would still be good to be able to add additional dimension/coordinate names on the fly. I would like to move the further discussion about this to #52.

from xmip.