Summary

All six tests properly work for a high number of points (N = 50000);
Properties comparison:
- Median estimation decreases in significance as N decreases;
- KDE for FWHM is the least stable test (not working for N \leq 1000);
- HSM swtill properly works for very few points (N = 200);
KS still properly works for very few points (N = 200);
Trapani test (less informative) decreases in significance as N decreases.

Questions

Bibliography