Details
-
Type:
Bug
-
Status: Done
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: ip_isr
-
Labels:None
-
Story Points:1
-
Team:Data Release Production
Description
Merlin Fisher-Levine was just puzzled as to why it didn't look like the interpolation of saturated pixels had been turned off despite having set doSaturationInterpolation=False. To stop the interpolation from ocurring, he had to set doSaturation=False. The docs say:
doSaturation = pexConfig.Field( |
dtype=bool, |
doc="Mask saturated pixels?", |
default=True, |
)
|
|
doSaturationInterpolation = pexConfig.Field( |
dtype=bool, |
doc="Perform interpolation over pixels masked as saturated?", |
default=True, |
)
|
indicating some false logic in the task. It looks like the fix should happen with a check for doSaturationInterpolation here: https://github.com/lsst/ip_isr/blob/master/python/lsst/ip/isr/isrTask.py#L1216
If these configs are meant to operate entirely independently, (a perhaps unlikely, but possible use case: the run() method was passed an image that had the SAT bit set elsewhere), this should be indicated clearly in the docs to avoid potential confusion along the lines "why are/aren't saturated pixels getting interpolated if I set doSaturation=False/doSaturationInterpolation=True).
Attachments
Issue Links
- verified by
-
DM-15683 Add tests for ip_isr/isrTask
- Done
Just to be aware of it. If there's no good place fo a task-level test now, Christopher Waters knows where it should go in
DM-15683.