Currently, we just have a single hard coded input-output testcase for the attitude filter. We should focus on some of the points below to make more/improved testcases:
Generate tests that run the filter for multiple iterations in a row
Potentially code a simple cxx sim to run the filter in with CI or create arrays of simulation data to feed the filter during CI and ensure convergence etc is happening as expected.
Testcases with a more complex initial state covariance (right now it's simply a diagonal matrix).
Currently, we just have a single hard coded input-output testcase for the attitude filter. We should focus on some of the points below to make more/improved testcases: