Clipping magnitude and phase #638
Open
yurirtonin wants to merge 9 commits intodevfrom
Open
Conversation
added 6 commits
September 15, 2025 09:42
- Changed clipping for both magnitude and phase, not only magnitude - Added clipping functionality to stochastic engines (CPU and cupy) - Renamed .cu file to clip_object - Renamed functions that referred to magnitude clipping
- Added soft clipping to ML algorithm (CPU and cupy) - Added it to stochastic and projectional serial engines - Small cleanups
Collaborator
Author
|
Fixed a couple of refactoring I had missed. Nonetheless, the latest test error on WASP_serial regarding the complex type of the object I am not sure how to solve yet. Running the WASP and WASP_cupu engines directly do not present this error |
daurer
reviewed
Sep 29, 2025
…ion to use the obj container instead of npy array
Contributor
|
Merge, but keep an issue for the implementtation of different strategies for amplitude + phase clipping (including the graph) |
Contributor
|
Add warning for not-availablilty in pycuda engines |
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The clipping input is now assumed to be either a 2-tuple or 4-tuple.
In case it is as 2-tuple, it is assumed to be (mininum_magnitude, maximum_magnitude)
In case it is as 4-tuple, it is assumed to be (mininum_magnitude, maximum_magnitude, minimum_phase,maximum_phase)
For ignoring a specific clipping, use None. For instance: (minimum_magnitude, None, None, maximum_phase)
I kept the approach that was already implemented for the magnitude clipping, so it is done right after object update and before probe update.
Done:
Still possible ToDos:
One open point of discussion: what is the "correct" way of doing such a clipping? Currently, we clip the magnitudes (np.abs) followed by clipping of the phase (np.angle), which would result in a behavior in the complex plane as the following: