Details
Description
The lsst.ip.isr.isrFunctions.overscanCorrection subtracts the overscan fit from the active image. It should also subtract the same fit from the overscan, allowing downstream processing to operate on the overscan-subtracted overscan (e.g., to measure the noise).
Attachments
Issue Links
- is duplicated by
-
DM-11266 Add option to ISR to estimate effective read noise from overscan region
- Invalid
Activity
Meredith Rawls, would you mind reviewing these changes, please?
price@pap-laptop-2:~/LSST/ip_isr (tickets/DM-14182=) $ git sub
|
commit 4f1a728bcb0529544310e57cba8806bf81d5f3be
|
Author: Paul Price <price@astro.princeton.edu>
|
Date: Mon Apr 23 14:03:49 2018 -0400
|
|
|
remove python 2 support
|
|
We no longer support Python 2.
|
|
|
python/lsst/ip/isr/__init__.py | 2 --
|
python/lsst/ip/isr/assembleCcdTask.py | 1 -
|
python/lsst/ip/isr/crosstalk.py | 1 -
|
python/lsst/ip/isr/fringe.py | 5 -----
|
python/lsst/ip/isr/isrFunctions.py | 3 ---
|
python/lsst/ip/isr/isrTask.py | 3 ---
|
python/lsst/ip/isr/linearize.py | 6 +-----
|
python/lsst/ip/isr/measureCrosstalk.py | 4 ----
|
8 files changed, 1 insertion(+), 24 deletions(-)
|
|
|
commit 2c715da74c6c2f7359b559af664831c2803ee6b1
|
Author: Paul Price <price@astro.princeton.edu>
|
Date: Mon Apr 23 15:07:53 2018 -0400
|
|
|
overscanCorrection: subtract fit from overscan
|
|
This allows downstream processing to inspect the fit results.
|
|
Because the overscan is now subtracted explicitly, this requires that the
|
overscan passed to this function not be contained within the image being
|
corrected (or the correction is applied twice). Unfortunately, it's not
|
clear how to detect this case; however, the primary user is IsrTask, which
|
pulls out the components explicitly and so shouldn't mix them.
|
|
Since this involved changes to the docstring, reformatted the docstring
|
for this function into numpydoc.
|
|
|
python/lsst/ip/isr/isrFunctions.py | 68 ++++++++++++++++++++++++++++----------
|
python/lsst/ip/isr/isrTask.py | 2 +-
|
tests/test_overscanCorrection.py | 20 ++++++++---
|
3 files changed, 68 insertions(+), 22 deletions(-)
|
|
|
commit e017d005828e9c695cb54f2371085661eb72614e (HEAD -> tickets/DM-14182, origin/tickets/DM-14182)
|
Author: Paul Price <price@astro.princeton.edu>
|
Date: Tue Apr 24 13:20:59 2018 -0400
|
|
|
IsrTask: add option to set variance from empirical read noise
|
|
The read noise is measured from the overscan.
|
|
This might be useful when working on detectors whose read noise hasn't
|
been measured, or drifts due to hardware tests.
|
|
|
python/lsst/ip/isr/isrFunctions.py | 2 +-
|
python/lsst/ip/isr/isrTask.py | 77 +++++++++++++++----
|
tests/test_empiricalVariance.py | 149 +++++++++++++++++++++++++++++++++++++
|
3 files changed, 212 insertions(+), 16 deletions(-)
|
Paul Price
added a comment - Meredith Rawls , would you mind reviewing these changes, please?
price@pap-laptop-2:~/LSST/ip_isr (tickets/DM-14182=) $ git sub
commit 4f1a728bcb0529544310e57cba8806bf81d5f3be
Author: Paul Price <price@astro.princeton.edu>
Date: Mon Apr 23 14:03:49 2018 -0400
remove python 2 support
We no longer support Python 2.
python/lsst/ip/isr/__init__.py | 2 --
python/lsst/ip/isr/assembleCcdTask.py | 1 -
python/lsst/ip/isr/crosstalk.py | 1 -
python/lsst/ip/isr/fringe.py | 5 -----
python/lsst/ip/isr/isrFunctions.py | 3 ---
python/lsst/ip/isr/isrTask.py | 3 ---
python/lsst/ip/isr/linearize.py | 6 +-----
python/lsst/ip/isr/measureCrosstalk.py | 4 ----
8 files changed, 1 insertion(+), 24 deletions(-)
commit 2c715da74c6c2f7359b559af664831c2803ee6b1
Author: Paul Price <price@astro.princeton.edu>
Date: Mon Apr 23 15:07:53 2018 -0400
overscanCorrection: subtract fit from overscan
This allows downstream processing to inspect the fit results.
Because the overscan is now subtracted explicitly, this requires that the
overscan passed to this function not be contained within the image being
corrected (or the correction is applied twice). Unfortunately, it's not
clear how to detect this case; however, the primary user is IsrTask, which
pulls out the components explicitly and so shouldn't mix them.
Since this involved changes to the docstring, reformatted the docstring
for this function into numpydoc.
python/lsst/ip/isr/isrFunctions.py | 68 ++++++++++++++++++++++++++++----------
python/lsst/ip/isr/isrTask.py | 2 +-
tests/test_overscanCorrection.py | 20 ++++++++---
3 files changed, 68 insertions(+), 22 deletions(-)
commit e017d005828e9c695cb54f2371085661eb72614e (HEAD -> tickets/DM-14182, origin/tickets/DM-14182)
Author: Paul Price <price@astro.princeton.edu>
Date: Tue Apr 24 13:20:59 2018 -0400
IsrTask: add option to set variance from empirical read noise
The read noise is measured from the overscan.
This might be useful when working on detectors whose read noise hasn't
been measured, or drifts due to hardware tests.
python/lsst/ip/isr/isrFunctions.py | 2 +-
python/lsst/ip/isr/isrTask.py | 77 +++++++++++++++----
tests/test_empiricalVariance.py | 149 +++++++++++++++++++++++++++++++++++++
3 files changed, 212 insertions(+), 16 deletions(-)
This is a good thing to add! Please see my comments over on GitHub.
Two main notes/questions: it's not clear to me how the read noise is usually set for the amplifier (is it ever based on reality?), and I do think it's important to (somehow) check that the overscanImage is not a sub-image of the ampMaskedImage since that is an easy thing to overlook if one were to pick up and use our overscan correction from scratch.
Jenkins is passing.
Paul Price
added a comment - Jenkins is passing. Thanks Meredith Rawls. I've posted answers on the GitHub PR. I believe this is ready to merge. Please let me know if you have any concerns. In the mean time, I'm running Jenkins on ip_isr to verify the small changes to the test haven't broken things.
Paul Price
added a comment - Thanks Meredith Rawls . I've posted answers on the GitHub PR. I believe this is ready to merge. Please let me know if you have any concerns. In the mean time, I'm running Jenkins on ip_isr to verify the small changes to the test haven't broken things. At Russell Owen's request, I am dropping the commit removing py2 support. He has more extensive changes on DM-14204 that I don't want to get in the way of.
Paul Price
added a comment - At Russell Owen 's request, I am dropping the commit removing py2 support. He has more extensive changes on DM-14204 that I don't want to get in the way of. Jenkins passed.
Paul Price
added a comment - Jenkins passed. Merged to master.
Paul Price
added a comment - Merged to master. 
Based on PFS Jira ticket PIPE2D-278.