cutTree

object of a class <code>factorMerger</code>

factorMerger

statistic used in the bottom-up search. Available statistics are:
<code>"loglikelihood"</code>, <code>"p-value"</code>, <code>"GIC"</code>.

stat

cut threshold or penalty (for GIC)

value

Splits factor levels into non-overlapping clusters
based on a <code>factorMerger</code> object.
If a <code>stat</code> is <code>"loglikelihood"</code> or "p-value"
then performs bottom-up search through models
on the merging path until spots a model scored worse
than the given threshold (<code>value</code>).
If <code>stat = "GIC"</code>, <code>value</code> is interpreted as
GIC penalty and optimal GIC model is returned..

The Merging Path Plot is a methodology for adaptive fusing of k-groups
with likelihood-based model selection. This package contains tools for
exploration and visualization of k-group dissimilarities.
Comparison of k-groups is one of the most important issues
in exploratory analyses and it has zillions of applications.
The traditional approach is to use pairwise post hoc tests
in order to verify which groups differ significantly. However,
this approach fails with a large number of groups in both interpretation
and visualization layer.
The Merging Path Plot solves this problem by using an easy-to-understand
description of dissimilarity among groups based on Likelihood Ratio Test (LRT) statistic (Sitko, Biecek 2017) <arXiv:1709.04412>.
'factorMerger' is a part of the 'DrWhy.AI' universe (Biecek 2018) <arXiv:1806.08915>.
Work on this package was financially supported by the 'NCN Opus grant 2016/21/B/ST6/02176'.

cutTree: Cut a Factor Merger Tree

Description

Usage

Arguments

Value

Details