RAD-158: Changed image units. #389
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #389 +/- ##
=======================================
Coverage 95.38% 95.38%
=======================================
Files 4 4
Lines 195 195
=======================================
Hits 186 186
Misses 9 9 ☔ View full report in Codecov by Sentry. |
CHANGES.rst
Outdated
| @@ -16,6 +16,8 @@ | |||
|
|
|||
| - Updated product table names. [#382] | |||
|
|
|||
| - Changed image units. [#389] | |||
There was a problem hiding this comment.
It may be good for the release notes to specify what the change it.
| @@ -26,9 +26,10 @@ properties: | |||
| tag: asdf://stsci.edu/datamodels/roman/tags/source_detection-1.0.0 | |||
| required: [photometry, wcs] | |||
| data: | |||
| title: Science Data (electrons / s) | |||
| title: Science Data (DN / s) or (MJy / sr) | |||
There was a problem hiding this comment.
Is the conversion to MJy/sr applied in Level 2 or L3? If L2 (and I may be confused about this) then remove MJy/sr
There was a problem hiding this comment.
L2, and the MJy/sr option was specifically requested in the ticket.
| oneOf: | ||
| - tag: tag:stsci.edu:asdf/unit/unit-1.* | ||
| - tag: tag:astropy.org:astropy/units/unit-1.* | ||
| enum: ["DN / s", "MJy.sr**-1"] |
There was a problem hiding this comment.
@braingram Can one of the tags be dropped here? I believe DN/s is a VO unit and supported by astropy.
There was a problem hiding this comment.
Thanks for the ping!
The oneOf here is needed as asdf-astropy will select which tag to use based on the unit (and if it is or is not a VO unit). See the relevant converter code here.
Writing out a ImageModel with DN / s data produces the following and uses the astropy tag:
data: !unit/quantity-1.1.0
unit: !<tag:astropy.org:astropy/units/unit-1.0.0> DN / s
value: !core/ndarray-1.0.0
source: 9
datatype: float32
byteorder: little
shape: [4088, 4088]whereas using MJy.sr**-1 produces the following and uses the stsci tag:
data: !unit/quantity-1.1.0
unit: !unit/unit-1.0.0 MJy.sr**-1
value: !core/ndarray-1.0.0
source: 9
datatype: float32
byteorder: little
shape: [4088, 4088]There was a problem hiding this comment.
Huh! My bad! I checked DN/s but not MJy/sr as I thought that surely this is a VO unit. Apparently not!
Resolves RAD-158
Closes #388
This PR changes the units of the variables within the Image Model schema.
NOTE: no regression test link, because these changes require update to regression test files.
Checklist
CHANGES.rstunder the corresponding subsection