_bearings_distribution: bin_centers terminology

[dhimmel]

This PR addresses terminology in internal-use functions for greater clarity:

• bin centers were referred to as "edges"
• bin_counts and bin_centers now have the same length, which is best for most use cases

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.98%. Comparing base (016d3ca) to head (4139a04).

Additional details and impacted files

@@           Coverage Diff           @@
##               v2    #1149   +/-   ##
=======================================
  Coverage   97.98%   97.98%           
=======================================
  Files          24       24           
  Lines        2436     2436           
=======================================
  Hits         2387     2387           
  Misses         49       49           

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

[gboeing]

@dhimmel help me understand what's going on here. Aside from the renaming changes, I see a couple numerical and indexing changes. What's the goal and what are they doing?

[dhimmel]

I see a couple numerical and indexing changes. What's the goal and what are they doing?

Currently, _bearings_distribution returns (bin_counts, bin_edges) where bin_edges are actually the centers of each bin with the first bin occurring at index position 0 and -1 (because 0 degrees is equivalent to 360 degrees).

The non-renaming changes are entirely related to removing the final repetitive value from the bin centers array. When giving the center of each bin, there is not much rationale for repeating the first bin at the end of the list. The repeated final value was a relic of numpy.histogram's more general design, which in the case of a polar / degree distribution is not needed.

When returning bin_centers, it is nice to have bin_counts be the same length.

[gboeing]

It looks like this comment should be updated then.

[gboeing]

Another reference to edges here.

[dhimmel]

Given the elegant but potentially confusing implementation of _bearings_distribution with respect to the bin splitting and merging, I revised variable names in this function as well as comments to provide greater clarity. Also added tests to test_bearings