Closed rongpu closed 2 years ago
Thanks for catching this @rongpu. Your changes look correct to me.
Can you please update "No changes yet" in the changes.rst file to describe this PR? I'll then merge this.
I'm working on a separate PR related to the randoms, which I should finish this week. Once I'm done with that, I guess I should make new versions of the randoms.
Thanks @geordie666 . I have updated the changes.rst file. Please feel free to make any edits.
As for producing new randoms, there is a separate issue with the WISE depth discrepancy, and I guess we might want to wait a bit in case we make any decisions/changes on that issue soon.
Yes, good point about waiting on the WISE-depth-issue. I guess the current randoms are "good enough" for most analyses, so we may as well wait.
I'll merge this PR once tests pass.
I think the unit test failure is unrelated to the updates in this branch, but I can't easily work on a fix until Cori is back. So, I'll merge this tomorrow morning and look at the unit tests then.
This PR fixes a bug in randoms.py that inaccurately rounds the pixel coordates. Specifically,
astype("int")
rounds towards zero, but it should be rounded to the nearest integer.