Feat: long tracker

[cpaniaguam]

This pull request includes several changes to the IceFloeTracker project, focusing on improving the tracking of ice floes over multiple days, adding new functionality, and updating the corresponding tests. The most important changes include the addition of the long_tracker function, updates to the notebook, and the introduction of new helper functions.

New functionality:

• Added long_tracker function to track ice floes over multiple days, including the necessary helper functions such as find_floe_matches, get_unmatched, get_trajectory_heads, and others. (src/tracker/long_tracker.jl, src/tracker/tracker-funcs.jl)
• Included the new long_tracker file in the main module. (src/IceFloeTracker.jl)

Notebook updates:

• Updated the track-floes.ipynb notebook to use the new long_tracker function and added new data processing steps and visualizations. (notebooks/track-floes/track-floes.ipynb)

Testing improvements:

• Added new tests for the long_tracker function, including cases with gaps in data. (test/test-long-tracker.jl)
• Updated test setup to use the new @ntestset macro for better test organization. (test/runtests.jl, test/test-matchcorr.jl)

Notes about long_tracker

1. This function can replace the call to _pairfloes within pairfloes which is used the pipeline workflow.

IceFloeTracker.jl/src/tracker/tracker.jl

Lines 226 to 246 in e7be057

	function pairfloes(
	segmented_imgs::Vector{<:FloeLabelsImage},
	props::Vector{DataFrame},
	passtimes::Vector{DateTime},
	latlonrefimage::AbstractString,
	condition_thresholds,
	mc_thresholds,
	)
	_pairs = _pairfloes(
	segmented_imgs,
	props,
	passtimes,
	condition_thresholds,
	mc_thresholds,
	)
	addlatlon!(_pairs, latlonrefimage)
	cols = [:ID, :passtime, :area, :convex_area, :major_axis_length, :minor_axis_length, :orientation, :perimeter, :latitude, :longitude, :x, :y, :area_mismatch, :corr]
	_pairs = _pairs[:, cols]
	return _pairs
	end

2. It has a simpler signature. _pairfloes as currently defined does too much work. Besides the matching, it also adds unique ids to floes, pass times, floe masks, and psi-s curves. long_tracker expects all those data to be included in the input props object. See sample workflows in the test suite and the illustrative notebook.

TODO

• Add tie-breaking criteria for potential matches.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[hollandjg]

Code looks good! I've a few clarifying questions. At this point, I'd really like to see some extra test images and trajectories with demonstrations that the tracking works over multiple days with gaps in-between.