correct_output.Rd

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/verify_mclust.R
\name{correct_output}
\alias{correct_output}
\title{Correct Output of Mclust}
\usage{
correct_output(mclust_result, min_proportion = NULL, simple = T)
}
\arguments{
\item{mclust_result}{Object of class Mclust (result of Mclust)}

\item{min_proportion}{Double - Minimum proportion a cluster should have}
}
\value{

}
\description{
This function checks if the output of a mclust restult is correct.
}
\details{
When using mclust to get the classification via mixture models wrong class assignments may occur. Meaning when the data points are not well seperated that multiple modeled probability distributions can overlap. For example when having a range
of data points between 0 to 10 and two exisiting distributions. The simplest assumtion in this case would be that all values from  0 >= 5 belong to the first classification cluster and values > 5 to the second one. When using the option modelNames = "V"
in Mclust to allow for varying variance, it happens that maybe values from 0-1 are assigned to cluster 2, values from >1 - 5 are assigned to cluster 1, and the remaining values are assigned to cluster 2 again. This is wrong and therefore one needs to check them
correct output.
}