Closed jacobdparker closed 7 months ago
Attention: 4 lines
in your changes are missing coverage. Please review.
Comparison is base (
93e06a2
) 60.18% compared to head (1753e61
) 63.40%.
Files | Patch % | Lines |
---|---|---|
mocksipipeline/inversion/response_matrix.py | 20.00% | 4 Missing :warning: |
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Looks like the tests are failing because the channel classes are instantiated differently now. Should just be a matter of fixing the tests to instantiate them correctly with an OpticalDesign
and a filter choice.
Alright I've reorganized things by renaming detector
to instrument
and moving the optical design stuff to optics
. Anything that gets instantiated gets put in a configuration.py
file.
Now that I'm seeing the length of one of these instrument configurations, I wonder if it makes sense to actually make instrument.configuration
a subpackage and then put each configuration in its own file. The optics configurations could stay as they are since they're relatively short.
Yeah, I can see that. It is very clunky that they are so long, but I do think having the math there is useful as a record. Otherwise they just seem like magic numbers.
Introducing two new classes
InstrumentDesign
andOpticalDesign
which make it easier to store configuration defaults and different MOXSI design iterations.