Add 3D plotting option to Plotterly

[mharris-dstl]

This PR gives the option of plotting in 3D using Plotterly. It also adds more tests and makes dimension checking in Plotterly more holistic. More details are:

• Plotting is now supported in 3D for ground truth, measurements, and tracks. Sensor plotting is not here but no reason why it can't be added in the future.
• Dimension checking has been simplified so that integers can always be used in cases like if self.dimension == 1:
• Track uncertainty has been added in the way of error bars in the x, y, and z direction, as suggested by @Carlson-J here. It is not perfect because it does not show times where covariance is large, but it's a simple way of getting an idea of track uncertainty. It makes use of scatterplot3d's error arguments - making custom error bars that plot the eigenvectors would require significant effort to implement. Unfortunately, I think a bug in Plotly means 3d error bars are fixed at 1 pixel wide, so they are quite hard to see. I think it's still better than not having them, but the functionality can be removed from this PR if desired. See the pictures below:

[gawebb-dstl]

Code changes look good from reading them. I'll try using the new code soon

[codecov]

Codecov Report

Attention: 11 lines in your changes are missing coverage. Please review.

Comparison is base (0a0f43b) 93.23% compared to head (6e8eadb) 93.40%.
Report is 23 commits behind head on main.

Files	Patch %	Lines
stonesoup/plotter.py	79.24%	3 Missing and 8 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #932      +/-   ##
==========================================
+ Coverage   93.23%   93.40%   +0.17%     
==========================================
  Files         200      200              
  Lines       12516    12538      +22     
  Branches     2580     2591      +11     
==========================================
+ Hits        11669    11711      +42     
+ Misses        595      585      -10     
+ Partials      252      242      -10     

Flag	Coverage Δ
integration	67.03% <22.64%> (-0.03%)	⬇️
unittests	88.84% <77.35%> (+0.25%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[spike-dstl]

Looks good and runs well. Couple comments that you may wish to think about but no issues. Looks good on the track-stitching example:

[spike-dstl]

The original logic allowed the number of mapping dimensions to be greater than the number of plotter dimensions. Both methods appear to be valid, and throwing an error rather than a warning is arguably more sensible. Nonetheless, I believe the following represents the original logic. Thought this was worth flagging, but feel free to mark as resolved.

Suggested change

	elif len(mapping) != self.dimension:
	raise TypeError("Plotter dimension is not same as the mapping dimension.")
	elif len(mapping) < self.dimension:
	raise TypeError("Plotter dimension is larger than the mapping dimension.")
	elif len(mapping) > self.dimension:
	warnings.warn("Mapping has length greater than plotter dimension. Additional mapping indices are ignored")

[csherman-dstl]

For consistency, does it make more sense to move the checks for particle and uncertainty within the "dimension == 2" block / move the other dimensions checks out here?

[mharris-dstl]

I was hesitant to do that because the dimension==2 line sits within a for-loop that goes through each track whereas uncertainty plotting is outside that. The current implementation plots all the tracks at once and then all the uncertainty at once. My worry is that changing the logic order to track, uncertainty for track, track2, uncertainty for track2, etc. may cause some unintended consequences. It's probably fine but I haven't tried it.

[gawebb-dstl]

All looks good to me. A good addition to the stone-soup plotting library.

You could change the plotting examples/MTT_3D_Platform.py to use Plotterly instead of Plotter.