-
Notifications
You must be signed in to change notification settings - Fork 65
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
Regression fix: if how='slice' or 'ray', we do _not_ want to warn #929
base: master
Are you sure you want to change the base?
Regression fix: if how='slice' or 'ray', we do _not_ want to warn #929
Conversation
…simply Does Not Work when I try it. Why?
I could use some feedback here @astrofrog @e-koch : I've removed warnings about dask loading whole cubes into memory, but I'm second-guessing myself now. When you do cube addition, what other option is there than to keep the whole cube in memory? On the other hand, should we just drop the warnings entirely? I am not sure they're ever appreciated or valuable. This is a bigger question, but for my own use, most of the time I do everything to suppress these warnings... |
… But, there's an open question about this.
… we have to turn off ignore_warnings
…getting rid of the test_arithmetic_warning.... why do we care for small cubes? we really don't
This passes completely if I run a local test:
but I reproduce the exact failure with tox. I don't understand it. |
For the cube addition: are we defaulting to a new memory map or zarr for the dask versions? If so, I don't see a particular issue with dropping the warnings. |
regarding the casa reading failures: I'm not sure the above is a 'real life' bug and not just something weird about the tox system. |
cube addition: I think it's a new in-memory array. But if we default to a new memmap, I'd be happier with that.... is that a thing we can do? We have to tell numpy to do it... |
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #929 +/- ##
==========================================
- Coverage 79.84% 79.79% -0.06%
==========================================
Files 24 23 -1
Lines 6024 6037 +13
==========================================
+ Hits 4810 4817 +7
- Misses 1214 1220 +6 ☔ View full report in Codecov by Sentry. |
Green on all but the CASA tests, for which it seems a STOKES keyword has gone missing. @preshanth, could you have a look? |
did a little investigating; it looks like this is a real issue, and we should check whether |
The issue seems to be that the cubes being used in the test have no stokes axis. These are perfectly valid cubes. You can see from the |
I did this and it gets over the error altogether.
|
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.
Looks good. One minor change to get rid of a couple print statements before merging.
@@ -107,8 +107,14 @@ def load_casa_image(filename, skipdata=False, memmap=True, | |||
print(desc['_keywords_']['coords']) |
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.
Remove temp print statements
#916 introduced a serious regression:
warn_slow
now completely ignoreshow='slice'
andhow='ray'
, which was not at all the desired behavior.