get.limits: Infer lower and upper mode position limits to define robust end-members.

Description

This function identifies the lower and upper limits within which robust end-members have clustered mode positions. It uses a kernel density estimate of the mode positions of all input end-member loadings, clips it at a user-defined minimum density and returns the resulting rising and falling shoulders of the kde peaks as limits.

Usage

get.limits(loadings, classunits, bw, threshold = 0.7)

Value

Numeric matrix with lower and upper mode limits.

Arguments

loadings: R object, output of function model.EM.
classunits: Numeric vector, optional class units (e.g. micrometers or phi-units) of the same length as columns of X.
bw: Numeric scalar, bandwidth of the kernel, moved over the data set. If omitted, the default value of 1 used.
threshold: Numeric scalar, threshold quantile which is used to identify mode clusters. Only kde densities above this values are kept and used to derieve mode cluster limits.

Author

Michael Dietze, Elisabeth Dietze

Details

Note that the threshold above which a mode cluster is identified is an arbitrary, user-defined value and probably needs to be adjusted iteratively to get reasonable results. The default value may or may not be adequate!

Examples

Run this code


## load example data set
data(example_EMpot)

## infer mode cluster limits
limits <- get.limits(loadings = EMpot)

Run the code above in your browser using DataLab

The Learning Leader's Guide to AI Literacy