-
Notifications
You must be signed in to change notification settings - Fork 14
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
DM-46052: Updates to IsrTaskLSST based on testing on real LATISS and LSSTCam images. #345
Conversation
d81d96c
to
dea8b19
Compare
This adds a saturated column to ensure that the masking code works, although it does add a bit more noise to a few of the test comparisons.
…ure. This tests the functionality of the bad amp filtering code in the crosstalk and ensures that one bad amp doesn't spoil the whole bag.
This also tests that neighbors of saturated regions in parallel overscan are also masked, in an effort to remove the chance of "negative" bias columns.
The crosstalk is assumed to exist and is run in typical bootstrap runs.
In bootstrap mode, no gains will be applied, and the config validation checks explicitly. This greatly simplifies what "bootstrap" mode means for downstream processing.
4ccab2f
to
7d57287
Compare
I reviewed the changes, and I spent some time reviewing test_isrTaskLSST.py. I didn't find anything wrong, although I don't understand all of it well enough. Marking the ticket reviewed. |
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.
I reviewed the changes, and I spent some time reviewing test_isrTaskLSST.py. I didn't find anything wrong, although I don't understand all of it well enough. Marking the ticket reviewed.
This PR contains a number of updates to ip_isr based on testing of the new task and pipelines from DM-45856. Aside from bug fixes and checks of invalid (nan) crosstalk matrix values, there are a large number of updates to the parallel overscan masking to ensure that (a) saturated flats don't get oversubtracted; (b) saturated and unsaturated defect bleeds into the parallel overscan region are handled correctly.
This also takes advantage of crosstalk
fitGains
to apply reasonable gain values when applying crosstalk corrections to bootstrap frames.