mixture
Whether the fitted density is a fair description of the sample it came from.
The question the rest of this report does not ask. mixtureAdequacy asks whether the sample can settle how many components there are; the criterion sections ask which count wins. All of them are about a structure this method recovers poorly. This asks whether the density is usable, which is what an input model is for, and it is the one section reporting on the output the method is built to get right.
The summaries come before the verdict deliberately. A test returns reject or do not reject. It does not say how far off the fit is on a quantity a simulation is sensitive to, and that is usually the question an input modeller actually has: a queue driven by these values responds to the mean and the coefficient of variation far more than to the shape between them, so an agreement of a percent on those may settle the matter whatever the test says.
Takes the assessment rather than computing it, for the reason mixtureReliability does: it refits the whole model on every replicate, and expensive work started by a call that looks like formatting is a surprise. The caller decides to pay for it.
Parameters
the assessment, from MixtureModeler.assessFitAdequacy
the section title