Pendulum has strange position and noise behavior at certain phases

[beckynevin]

Describe the bug
Try running at time = 1. I'm seeing strange behavior where the position is zero, even when I see that when I run an entire series the point at t = 1 is not at zero. The noise is also not normally distributed which is strange.

I'd like to test this bug by adding some tests that align a lot of individual times with the run from a time series.

This is my fault most likely.

[beckynevin]

I think some of this bug was addressed via issue #121 , but I'd still like to do some more detailed testing to verify this. A lot of this testing will be happening in deepuq, which relies upon a pretty in-depth use of the pendulum module.