-
-
Notifications
You must be signed in to change notification settings - Fork 128
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
What does compressed_axes = None
mean?
#535
Comments
There are a few cases where
I think that is makes sense to have a default value for the cases that users don't have a preference. I used On the one hand I don't think it's really an issue because any array with
On the other hand I can see how such behavior could be confusing. In terms of removing it, I think that it should stay for 0- and 1-dimensional arrays. For higher dimensional arrays, we could pick a different default value - something like |
I think this may mislead people, and make it hard to get at the underlying format. Would
This seems reasonable to me, as long as there is a concrete
As said above we could use either
I think |
Okay, that makes sense to me. I like |
Just bumped into this, there are cases e.g. in to_scipy_sparse that assume non-None compressed axes, but a GCXS array can still be constructed with |
I was going through the
GCXS
code trying to fix bugs when I came across a few places wherecompressed_axes
was being set toNone
. That, in my opinion, should never happen. It should always be a tuple of integers.@daletovar Could you specify the context here, and what could be needed to remove this?
The text was updated successfully, but these errors were encountered: