Nirspec improvements in blotting for outlier detection, JP-2050 #6326
Conversation
|
Looks like a couple of unit tests need updating (see CI failures) |
Codecov Report
@@ Coverage Diff @@
## master #6326 +/- ##
==========================================
- Coverage 78.00% 77.66% -0.35%
==========================================
Files 402 402
Lines 34633 34841 +208
==========================================
+ Hits 27017 27060 +43
- Misses 7616 7781 +165
*This pull request uses carry forward flags. Click here to find out more.
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Looks OK mechanically to me, although I'm no expert on IFU WCS transforms. My eyes are picking up what I'm guessing will be flake8 failures, but I don't see the results of that in the latest CI test run. It's only showing the results of the docs build. Should try to trigger a rerun of all CI tests.
jwst/cube_build/ifu_cube.py
Outdated
| @@ -221,7 +221,7 @@ def define_cubename(self): | |||
| fg_name = fg_name.lower() | |||
| newname = self.output_name_base + fg_name + '_s3d.fits' | |||
| if self.output_type == 'single': | |||
| newname = self.output_name_base + fg_name + '_single_s3d.fits' | |||
| newname = self.output_name_base + fg_name + '-single_s3d.fits' | |||
| if self.coord_system == 'internal_cal': | |||
| newname = self.output_name_base + fg_name + '_internal_s3d.fits' | |||
There was a problem hiding this comment.
If the "single" file names are being changed to use a hyphen as the separator, instead of underscore, should the "internal" names be changed the same way here?
d3e9ec9 to
5f6d6c4
Compare
|
@eslavich @hbushouse I have been getting the CI/macOS failure quite a bit recently. Is there really something wrong or do I just ignore it ? I don't understand the information when I click on 'Details' |
|
I've been running into macOS errors as well, but they don't seem to be consistent (I've seen a failure and a success after rerunning the same commit). The doc build failure can be resolved by rebasing on master. |
|
@eslavich |
5f6d6c4 to
beb27a5
Compare
That only happens if there's a conflict to resolve. It sounds like git was able to rebase automatically, which means all you need to do afterwards is force push to the remote. |
|
@hbushouse merged code to hopefully fix NIRSpec blotting. So you can continue with testing |
…etelescope#6326) * updates * updated in_ifu_slice * upates * update test using in_ifu_test * flake8 fix for test * Updated change log * fix single name * fix typo
Partially resolves JP-2050
Description
There was a bug in the NIRSpec IFU blotting. An new routine was added to assign_wcs.util.in_ifu_slice that determines which ra, dec, lambda of the median cube fall in a NIRSpec slice. blot_cube_cube.py was updated to use this information
Checklist