Open gmacgilchrist opened 1 year ago
Which way do you see this integration going? Importing doralite
into gfdl_utils
or vice versa?
I've found this integration do be incredibly efficient. I used to copy-paste PP paths from DORA and then manually explore them from the command line to see what output was available. Now this is as simple as:
import doralite
import gfdl_utils.core as gu
pp_name = gu.find_variable(pp, 'TKE_itidal')[0]
local = gu.get_local(pp, pp_name, "ts")
path = gu.get_pathspp(pp, pp_name, "ts", local, "*", "*")
Really great that you're making use of this @hdrake (especially now that I can't, since my GFDL access is revoked). I had envisioned integrating doralite
into gfdl_utils
, or at least just using them alongside one another. I just wanted to be able to use the o-div
number to get the pp path (or any other element that doralite
can retrieve), and then use gfdl_utils
from there.
The
doralite
(repo here) package leverages some of thedora
functionality to determine model metadata (for example the pp directory path) within python. It is now available to be used on PP/AN. It makes sense to integrate this with the functionality ofgfdl_utils
, so that one can go seamlessly from, say, the o-div number of an experiment, todmget
-ing and opening specific variables.