Details
-
Type:
Bug
-
Status: Invalid
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: ip_diffim
-
Labels:None
-
Story Points:0
-
Epic Link:
-
Sprint:AP S20-2 (January), AP S20-3 (February), AP S20-4 (March), AP S20-5 (April), AP S20-6 (May), AP F20-1 (June), AP F20-2 (July), AP F20-5 (October), AP F20-6 (November), AP S21-3 (February), AP S21-4 (March), AP S21-6 (May)
-
Team:Alert Production
Description
The decorrelation afterburner either performs very little under- or very large overcorrection after changing the order of the images in DM-21809 in the science convolution case. Perhaps this is true in general, just was not apparent in the "normal" template convolution case given the negligible noise in the template.
Hints:
- Check whether there is an assumption regarding the normalization of the AL solution PSF.
- Check the Fourier space operations in the implementation. There may be padding and low-/high frequency region swapping issues.
Attachments
Issue Links
- is duplicated by
-
DM-9441 Additional debugging of ZOGY and A&L(decorrelation)
- Invalid
-
DM-10954 Investigate optimizing mask setting in ImageMapReduce
- Invalid
- is parent task of
-
DM-25050 Fix non-normalized matching kernel case in the decorrelation afterburner
- Done
-
DM-23854 Add spatially varying PSF solution to fixed decorrelation afterburner correction
- To Do
-
DM-25051 Implement proper variance plane calculation in the decorrelation afterburner
- To Do
-
DM-21920 Try out and write notebook about demonstrating basic DFT space steps for ip_diffim operations
- Done
-
DM-24371 Implement fixed correction fixed PSF support decorrelation afterburner
- Done
-
DM-24316 Demonstrate noise whitening and numerical precision in frequency space
- Won't Fix
-
DM-23857 Add support for spatially varying correction to the decorrelation afterburner
- Invalid
- relates to
-
DM-23855 Audit and debug current zogy image differencing code
- Done
-
DM-23853 Create test cases for the decorrelation afterburner and image differencing
- In Progress
-
DM-21809 Check the variance plane operations and the decorrelation afterburner operations in the science convolution case
- Done
Gabor Kovacs [X] You may want to dig through https://jira.lsstcorp.org/browse/DM-6580 to see if there is any relevance.