perturb

Perturb data by adding Gaussian noise, bootstrap resampling or mix-up. Gaussian noise has mean 0 and variance 0.01*average variance of all variables. The mix-up lambda is 0.9.

Providing mean partition for ensemble clustering by optimal transport alignment(OTA), uncertainty measures for both partition-wise and cluster-wise assessment and multiple visualization functions to show uncertainty, for instance, membership heat map and plot of covering point set. A partition refers to an overall clustering result. Jia Li, Beomseok Seo, and Lin Lin (2019) <doi:10.1002/sam.11418>. Lixiang Zhang, Lin Lin, and Jia Li (2020) <doi:10.1093/bioinformatics/btaa165>.

Lixiang Zhang

OTclust

Mean Partition, Uncertainty Assessment, Cluster Validation and
Visualization Selection for Cluster Analysis

Beomseok Seo

Lin Lin

Jia Li

perturb function

<dl><dt>data</dt>
<dd>-- data that will be perturbed.</dd>
<dt>method</dt>
<dd>-- adding noise is <code>0</code>, bootstrapping is <code>1</code> and mix-up is <code>2</code>. Default is adding noise.</dd></dl>

Arguments

Perturb data by adding noise, bootstrapping or mix-up — perturb

<dl>

<dt>data</dt>
<dd>-- data that will be perturbed.</dd>


<dt>method</dt>
<dd>-- adding noise is <code>0</code>, bootstrapping is <code>1</code> and mix-up is <code>2</code>. Default is adding noise.</dd>

</dl>

perturb: Perturb data by adding noise, bootstrapping or mix-up

Description

Usage

Value

Arguments

Examples