Closed igoumiri closed 1 year ago
Tests are failing but kind of randomly.
This actually explains something odd Alec and I noticed last week, but I didn't realize the problem was in the CI chassis.
Tests are failing but kind of randomly.
The tests that are failing do so because they are optimizing while we check to see if the output is within some tolerance of the desired answer. There is some randomization, so if we sample a bad batch or something the error can be over the tolerance. We might need to increase the tolerances, since we run these tests so many times.
I think you can also guard against the multivariate error by changing
to
self.assertLessEqual(mse, 0.25)
Fixes https://github.com/LLNL/MuyGPyS/issues/150