Radii estimation

[sidneymau]

Currently, the simple binner uses a statistical approach to estimating the galactic radii and records these values; these radii are used for the quoted significance value. However, as these estimates are often very poor, the plot renderer makes a data-driven estimate of the radii and uses these different values for all of the plots. This makes a disparity between what is plotted and what is used for the significance computation.

This could be resolved by choosing either radius consistently.

• The statistical radius is preferred because it finds the radius of maximum physical significance; however, this is often a characteristically unreal estimation.
• The data-driven radius is preferred because it prefers a realistic radial profile.

Alternatively, there are a few (potentially) better options:

1. Use the statistical radius to locate hotspots and then use the data-driven radius to recompute the statistical significance and make the plots.
2. Improve the statistical radius (somehow...) such that it returns realistic radii estimates comparable to the data-driven radii.