Factor orbit, FOV, and observing constraints into new Mission class

[lpsinger]

I am not sure if this is the right abstraction, but at least it makes it so that the Dorado observing constraints are not in the one global visibility_constraints list.

@mcoughlin, could you see using this new Mission class?

[codecov[bot]]

Codecov Report

Merging #40 (99340c1) into master (da6db2f) will increase coverage by 1.21%. The diff coverage is 77.77%.

@@            Coverage Diff             @@
##           master      #40      +/-   ##
==========================================
+ Coverage   23.52%   24.73%   +1.21%     
==========================================
  Files          25       26       +1     
  Lines        1424     1435      +11     
==========================================
+ Hits          335      355      +20     
+ Misses       1089     1080       -9     

Impacted Files	Coverage Δ
dorado/scheduling/scripts/animate.py	8.25% <16.66%> (-0.44%)	:arrow_down:
dorado/scheduling/scripts/main.py	9.61% <33.33%> (-0.39%)	:arrow_down:
dorado/scheduling/mission.py	96.00% <96.00%> (ø)
dorado/scheduling/constraints/__init__.py	84.61% <100.00%> (-2.06%)	:arrow_down:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update da6db2f...99340c1. Read the comment docs.

[lpsinger]

Looks really nice. And I do like the python configurations.

This doesn't use the Astropy configuration system. This just introduces a new class.

Can you see using this in the UVEX scripts?

We should also enable turning roll on or off perhaps? For now, I was solving with no roll.

For now, since Dorado hasn't decided whether or not to vary roll angles, I am thinking of the roll grid as a command line option. If you pass a roll step size of 90 degrees, that has the same effect as turning off the roll grid.

[mcoughlin]

Yes. I suppose just slight annoyance if you want to update mission constraints needing to modify repository, but I think it's fine.

[lpsinger]

Yes. I suppose just slight annoyance if you want to update mission constraints needing to modify repository, but I think it's fine.

Fair point. I'll whip up an example of what the Astropy configuration system implementation would look like.

[lpsinger]

I couldn't get the Astropy configuration system to work. I think it has to do with our module being called dorado.scheduling rather than dorado_scheduling. Can we adopt the Mission class as simply an incremental improvement?

[mcoughlin]

definitely better than nothing!