mixture
fun ReportBuilder.mixtureWorthIt(results: MixtureModelingResults, title: String = "Was a mixture worth it?")(source)
The recommendation against the best single distribution.
The comparison is exact rather than approximate: a one-component mixture is a single distribution, drawn from the same catalog and scored by the same criterion, so no separate fit is involved. The goodness-of-fit tests are given for both because a criterion difference alone is on an arbitrary scale and says nothing about whether either fit is defensible.
Parameters
results
the modeling results
title
the section title