Mean motion based "propagation" for N objects in parallel

[ChristopherRabotin]

High level description

Given just the mean motion, it's possible to compute the the future state vector of an orbit.

Requirements

Given an Orbit structure, compute its future orbit given only two body dynamics of the current central body.

Test plans

TBD

Design

The equation is:

M = E - e*sin(E)

where M is the mean anomaly, E is the eccentric anomaly, and e is the eccentricity. Once you've solved for E, you can calculate the true anomaly (ν) using the following equation:

ν = 2*atan(sqrt((1+e)/(1-e))*tan(E/2))

Then use the from_keplerian initializer to return a new orbit.

This should also include a quick way to create a simple dummy frame with a demo GM value.

This should be available from Python and allow for quick computation of very many orbits at once. This should likely use rayon in Rust to spam the CPUs.

[ChristopherRabotin]

Done in #191.