-
Notifications
You must be signed in to change notification settings - Fork 20
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
Adding stats functions #694
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.
please rename the field
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.
Great work! This is looking good! I've added some comments about packaging. This PR needs some tests as well. You could add tests to interface_tests.jl. Be sure to leave a link to the github issue to document what precisely the tests are testing.
Let's also check that the shape of the lazy tensors is always equal to the shape of the returned tensors when in debug mode using an |
4054792
to
edcb1ca
Compare
@@ -111,7 +109,8 @@ function remove_reorders(prgm::LogicNode) | |||
if length(idxs2) < length(getfields(arg)) | |||
reorder( | |||
aggregate( | |||
nothing, nothing, arg, setdiff(getfields(arg), idxs2)... | |||
overwrite, nothing, arg, |
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.
@kylebd99, I found this was necessary to get things to work, and I'm not sure why. The issue was that merge_stats_def would try to evaluate the aggregate function because the original rule to remove isnothing(init)
aggregates never matched. I'm just confused about how we never hit this before.
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.
Hmm, yeah I don't have a lot of intuition on this one. It seems like the original version should have checked isnothing on op
not init
. But, I'm quite surprised that it was never firing.
Codecov ReportAttention: Patch coverage is
... and 4 files with indirect coverage changes 🚀 New features to boost your workflow:
|
fixes #574
Implementing size property to lazy tensors to use with stats functions.