[PIPE2D-881] Ring like residuals after PIPE2D-874 Created: 06/Aug/21  Updated: 26/Oct/21  Resolved: 26/Oct/21

Status: Done
Project: DRP 2-D Pipeline
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: ncaplar Assignee: ncaplar
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: PNG File example_defocus.png     PNG File example_focus.png    
Story Points: 6
Sprint: 2DDRP-2021 A 10
Reviewers: hassan

 Description   

Certain spots show ring like features in the residuals. This is due to the fact that the solution have not adapted to the new algorithm from PIPE2D-874. It seems that this is due to poor modelling of the struts (see defocused example.). See PIPE2D-700 for prevous example of similar problem.

 

 

 Comments   
Comment by ncaplar [ 08/Sep/21 ]

The reanalysis started today (sent to cluster on September 7) after the changes and improvements to the algorithm in PIPE2D-880 (internal version of Zernike_Module 0.48)

Comment by ncaplar [ 10/Sep/21 ]

First set of images and looks better. I have sent all the spots to be analyzed on the cluster. 

Comment by ncaplar [ 17/Sep/21 ]

The analysis has finished and new results have been produced at /tigress/ncaplar/PFS/Fit_Results/Sep0521/ . The ``only'' thing left is to integrate them in the pipeline.

Comment by ncaplar [ 07/Oct/21 ]

I do not believe that there is actually review to be made, as no code has been changed - still, I am happy to go through formal procedure, if nothing else to exercise good practices. 

Comment by hassan [ 15/Oct/21 ]

Changes look fine. No additional comments made for PR https://github.com/Subaru-PFS/drp_pfs_data/pull/15 .

Comment by rhl [ 15/Oct/21 ]

Reading this ticket I don't know what happened. Was there a problem? Is it fixed?

Comment by ncaplar [ 15/Oct/21 ]

Indeed, this is all very unclear as it has written here. The basic algorithm introduced in PIPE2D-874 was fine, but as I was rushing to create the solution in the previous ticket, I did very short runs where I was searching for new optimal parameters. As such, there were many locations the solution did not converge. I reran everything with longer running time and everything looks fine. So indeed there was no deep problem in the algorithm. That is partially why this ticket is poorly commented, as I did not give it much importance - but this has caused a lot of confusion.

Generated at Sat Feb 10 15:59:17 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.