Rather than performing tests of CubeLineMoment on actual data, it would be convenient to be able to perform tests on a synthetic image cube. Perhaps this could be an image cube generated with:
(1) A defined level of per-channel noise
(2) Several Gaussian (and may non-Gaussian) lines which possess spatial (but known) variability in width and intensity
Rather than performing tests of CubeLineMoment on actual data, it would be convenient to be able to perform tests on a synthetic image cube. Perhaps this could be an image cube generated with:
(1) A defined level of per-channel noise
(2) Several Gaussian (and may non-Gaussian) lines which possess spatial (but known) variability in width and intensity