New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add analog downscaling prototype #98
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is cool, @dgergel. I'm pumped to see how this scales.
I have a few inline comments and corrections — as you can see.
In addition:
- Add this change to HISTORY.rst. Cite this PR and include your github username.
- Could we squeeze in at least one
service.py
-test to cover new behaviors/services? I think it'd be fine moving/adapting thecore.py
tests to fill this gap, if you don't want to write something new (because if service.py is tested, then core.py is tested). But I think we need test coverage for those new services. - In
dodola/tests/test_cli.py
could you addtrain-aiqpd
andapply-aiqpd
to thetest_cli_helpflags
test. The IDs for this would betrain-aiqpd --help
andapply-aiqpd --help
, respectively.
@brews tests are passing, this is ready for another review! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks! I have a few other quick changes needed! I'll talk with you about setting up tests.
@brews spoke too soon. I need to make a couple updates now that we're pulling from my updated branch of |
@dgergel Okay! Appreciate the heads up! |
@brews okay now it's actually ready! |
For the record it looks like this is running with |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This has all the changes I requested. Thanks!
We're going to want to be sure to pin this new xclim
to a commit hash or tag before the next release.
This PR adds the analog-inspired, quantile-preserving downscaling method as a new service. It is split up into two services,
train_aiqpd
andadjust_aiqpd
and is based on the implementation of QDM. This is intended to supersede spatial disaggregation as the main mode of downscaling indodola
for the time being.Basic CLI interface is:
closes #73