Closed RuneDominik closed 1 year ago
I actually like the test with specific, artificially created test data better checking a specific result better. This is a much more stringent test than what you have now on the prod5b IRFs.
Keep both to assert it works on "real" data and give a more understandable test to future developers? While stringent, the old test is also unintuitive in its data creation and therefor not easy verifiable.
This PR changes the test for the aeff interpolation to use actual irfs instead of a dummy effective area.