Initial Python bindings

[ChristopherRabotin]

Status

Early draft, pull request is opened mostly to run through the test suite, closes #93 .

What's left for the next demo of Styx to run a whole simulated OD from a Python workflow.

• Build a Python interface to a TrackingArcSim<Spacecraft, StdMeasurement, GroundStation> to be able to generate measurements from Python and the proxy functions for generating the measurements (requires loading a trajectory)
• Build the bindings to initialize a Ground Station from the YAML (via the StationSerde which implements the Configurable trait)
• Build a structure to store the measurements into an in-memory and in-parquet file after calling generate_measurements in the simulated tracking arc
• Load such file and rebuild that structure from Rust and Python
• Update ODP to use the measurement structure (which is shared between simulated and real measurements)
• Export the new Traj<S> to parquet (useful for plotting and required for setting up a TrackingArcSim)
• Import a trajectory parquet file from Python

Then:

• Migrate Orbit and Spacecraft state definitions from Thomas' work (worked out of the box!)
• Implement Configurable for Orbit and Spacecraft to build each from a YAML
• Implement Configurable for propagator and dynamics (should be similar to current set up with Config.rs). I think it's much easier to set this up from YAML than from the Python directly. It's also easier to pass around between functions I think but I may be wrong here. @thomasantony if you have insights or a preference, let me know. The use-case I have in mind is to set up a mission design or orbit determination run just like one would set up a machine learning experiments following a "data engineering convention" approach
• Implement Configurable for OD process to set up an OD process from YAML & Python Instead, the OD process is set up manually, cf. tests/python/test_orbit_determination.py .
• Event searching in Python propagation

Soon thereafter:

• Add mission events to Traj<S>, which may be generated from a "find events" call (i.e. pass an event finding to a trajectory to store all of the found events directly as mission events, which is helpful for analysis). Consider implementing Allow Boolean combination of Events #97 which doesn't seem like too much work. Moved to Trajectory information with custom files to be exported as Parquet files instead of CSVs #120
• Implement Configurable for targeters Moved to Refactor optimizer/targeter considering ease use in Python #119

Effects

• Adds Python bindings to Nyx, huge thanks to @thomasantony who coded this up years ago and I didn't get around to merging it.
• Updates version to 2.0.0-alpha.0

If this change adds or modifies a validation case

• No. But it uncovered a bug in the rotation of vectors. This will be fixed in Use SPICE files via ANISE for import/export and computations #86