Repository navigation
Add test for real FFT of misaligned input arrays - #3087
Merged
Merged
Conversation
antonwolfy
requested review from
ndgrigorian and
vlad-perevezentsev
as code owners
October 5, 2026 10:19
Contributor
|
View rendered docs @ https://intelpython.github.io/dpnp/index.html |
Contributor
|
Array API standard conformance tests for dpnp=0.21.0dev12=np2py314h8d9cdd5_6 ran successfully. |
oneMKL float32 rounding residue in zero bins is coherently amplified by prod(shape) for constant input, exceeding the fixed atol=1e-3 inherited from CuPy on some CPUs. Use eps * prod(shape) instead.
Collaborator
|
coverage: 78.669%. remained the same — add-rfft-input-alignment-test into master |
vlad-perevezentsev
approved these changes
Oct 6, 2026
vlad-perevezentsev
left a comment
vlad-perevezentsev
left a comment
Contributor
There was a problem hiding this comment.
LGTM
Thank you @antonwolfy !
github-actions Bot
added a commit
that referenced
this pull request
Oct 6, 2026
This PR adds `test_rfft_input_alignment` to the third-party FFT tests. It covers real-to-complex FFTs whose input data pointer is not aligned to the size of a complex output element. The test runs `dpnp.fft.rfft` (1-D) and `dpnp.fft.rfftn` (3-D) on float32/float64 inputs (float64 only on devices that support it), built in three ways, each at element offset 0 (aligned) and 1 (misaligned): - a contiguous view into a larger array (`a[offset]`) - a strided view (`a[..., offset::2]`) - an array created over an offset buffer (`dpnp.ndarray(shape, dtype=dtype, buffer=backing, offset=offset)`) Each case checks that the result matches NumPy and that the backing buffer is left unchanged. These cases already work on Intel CPU and GPU devices, so no change to the implementation is needed; the test guards against regressions. e22aa78
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds
test_rfft_input_alignmentto the third-party FFT tests. It covers real-to-complex FFTs whose input data pointer is not aligned to the size of a complex output element.The test runs
dpnp.fft.rfft(1-D) anddpnp.fft.rfftn(3-D) on float32/float64 inputs (float64 only on devices that support it), built in three ways, each at element offset 0 (aligned) and 1 (misaligned):a[offset])a[..., offset::2])dpnp.ndarray(shape, dtype=dtype, buffer=backing, offset=offset))Each case checks that the result matches NumPy and that the backing buffer is left unchanged.
These cases already work on Intel CPU and GPU devices, so no change to the implementation is needed; the test guards against regressions.