To euler fix

[kcamroccm]

Fixes: #
Issue with to_eularian_angles() method where quaternions with pitches of -90 or 90 would not return the correct values. These pitches need to be handled differently

About

Handling the singularity that occurs when when the pitch is -90 or 90, to have to_eularian_angles() work for all rotations.

How Has This Been Tested?

I wrote several test scripts in another branch called to_euler_tests in my fork https://github.com/kcamroccm/Colosseum/tree/to_euler_tests/PythonClient/TMM

There is a pytest unit test in file test_to_euler.py that goes through random orientations with the singularity exactly, near the singularity, and fully random orientations, converts them back and forth using the utils.to_quaternion() and to_eularian_angles() and verifies that they are still equal within a threshold.

There is also a script called vision_pawn_rotation.py that I used to verify the conversion visually while connected to an to unreal. This also works with random orientation of the same types used the pytest script

Screenshots (if appropriate):