-
Notifications
You must be signed in to change notification settings - Fork 24
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
Fix edgecase with 1 input value that is not selected in subsample_data #114
Conversation
np.array([20, 21, 22, 23, 24]), | ||
torch.IntTensor([0, 1, 4]), | ||
), | ||
# Edge cases |
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.
can you be more specific one each cases ?
ex: "one value case" ..
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.
Minor changes in formulations (comment and changelog), but code looks good!
For reference here is a related issue explaining the cause of this edge case: #83 (comment)
], | ||
) | ||
def test_subsample_data(x, idx, choice): | ||
# points w. |
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.
If this is a dead comment, can it be removed?
CHANGELOG.md
Outdated
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.
Could you be a bit more explicit here, and specify that this fix is related to subsampling?
ef43f34
to
36236f4
Compare
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 for the comments
36236f4
to
31c7cb7
Compare
31c7cb7
to
0be5fa2
Compare
In the preprocessing steps when subsampling data, there was a buggy edge case when there was only one point in the batch that had to be removed.
Up to now, the point was not removed (side effect of trying to force the output to be a Tensor when only one value was kept)
@MichelDaab or @gliegard can you have a look as @CharlesGaydon is not available?
Thanks