crawlingcub

It seems reducing the `num_gauss_hermite_points` may sometimes cause the assertions to fail. I have updated the tolerances in the assertions to account for this behavior.

Adding @st-- please let us know if these changes look ok. Here I am assuming that getting `-inf` when computing log densities with high `num_gauss_hermite_points` is due to over-approximations as...

Hi @josejimenezluna , sorry i edited the wrong test. The "fit" function in `tStudentProcess.py` doesn't currently accept a trials argument. So we cant pass it from the test as well....

@josejimenezluna Just to add, if you have any suggestions about how to refactor the test in a different way, I will be happy to incorporate that. Thanks!

@googlebot I signed it!

Hi, I fetched and updated the branch. Can you merge into master?