-
Notifications
You must be signed in to change notification settings - Fork 1.4k
2474 Deprecate SegmentationSaver and TransformInverter handlers #2475
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
Conversation
merge master
merge master
merge master
Signed-off-by: Nic Ma <nma@nvidia.com>
|
/black |
Signed-off-by: monai-bot <monai.miccai2019@gmail.com>
|
Hi @ericspod , Thanks for developing the from monai.handlers.utils import evenly_divisible_all_gather
print(evenly_divisible_all_gather(["tests"]))It can print out the warning message if running in a regular single process program. Thanks. |
|
I think you're using incorrect arguments unless the ones you provided were for testing. The Also for the For the multiprocessing this relates to the mechanism I added to ensure the warning is show only once. A global dictionary stores definitions that were called/instantiated and have printed the warning, on subsequent calls/instantiations the warning is skipped. This won't necessarily work across processes if this dictionary wasn't filled in before spawning subprocesses, so each process will start with an empty dictionary and each should issue a warning. This will always happen in Windows which lacks fork semantics. If the stdout/stderr for subprocesses is suppressed or routed elsewhere you will never see the warning. However looking at the tests, for |
|
Hi @ericspod , Thanks for your detailed explanation! May I ask what the Thanks. |
Signed-off-by: Nic Ma <nma@nvidia.com>
|
If you use a |
I have updated the PR according to your comments to use the Thanks. |
|
Just a reminder that I pushed the tag |
|
Hi @wyli , Thanks for your help! I restarted the CI tests, let's see whether it can pass now as you pushed a new tag. Thanks. |
This should already be the behaviour. The |
|
Hi @ericspod , Thanks for your suggestions! Maybe I misunderstood your points. So is it correct in the PR now? @deprecated(since="0.6.0", removed="0.7.0", msg_suffix="XXX") |
|
Seems the CI test failed due to below error: Thanks. |
Yes that should be right. |
It doesn't like versions of the form "0.6.0rc0" because "0rc0" part won't parse. I had assumed that versions would always be three numbers separated by "." followed optionally by "." and other text that's ignored. A version of the form "0.6.0.rc0" would work otherwise we can figure out how to split the version string by "rc#" if it occurs. |
|
Hi @ericspod , Maybe Thanks. |
I hadn't want to rely on |
|
Cool! Thanks for your quick update! Thanks. |
|
PR #2482 should sort the issue now. |
|
Hi @ericspod , Cool! Thanks for your quick update. Thanks. |
ericspod
left a comment
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.
I'm good with it if @wyli is too.
wyli
left a comment
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, perhaps you could also add some docstring for things that were not clear to you?
Signed-off-by: Nic Ma <nma@nvidia.com>
|
Sure, enhanced the doc-strings to make it more clear. Thanks. |
|
/black |
|
looks like it's related to version comparison, I'm looking into this |
Can your PR #2485 fix it? Thanks. |
yes, tested #2485, it is good now. I didn't address #2485 (comment) but I think it's good enough for v0.6.. |
Fixes #2474 .
Description
This PR added
deprecateddecorator toSegmentationSaverandTransformInverterhandlers as we already have better options.Status
Ready
Types of changes
./runtests.sh -f -u --net --coverage../runtests.sh --quick --unittests.make htmlcommand in thedocs/folder.