compute_metric

Metric will be computed from 2 to max_k clusters. Note that the row number in results will be different from k.

A Shiny application and functions for visual exploration of hierarchical clustering with numeric datasets. Allows users to iterative set hyperparameters, select features and evaluate results through various plots and computation of evaluation criteria.

Rafael Henkin

visxhclust

A Shiny App for Visual Exploration of Hierarchical Clustering

compute_metric function

<dl><dt>dmat</dt>
<dd>distance matrix output of <code>compute_dmat()</code> or <code><a href="/link/stats%3A%3Adist()?package=visxhclust&version=1.1.0" data-mini-rdoc="visxhclust::stats::dist()">stats::dist()</a></code></dd>
<dt>clusters</dt>
<dd>output of <code>compute_clusters()</code> or <code>fastcluster::hclust()</code></dd>
<dt>metric_name</dt>
<dd>"silhouette" or "dunn"</dd>
<dt>max_k</dt>
<dd>maximum number of clusters to cut using <code>dendextend::cutree()</code>. Default is 14.</dd></dl>

Arguments

Compute an internal evaluation metric for clustered data — compute_metric

<dl>

<dt>dmat</dt>
<dd>distance matrix output of <code>compute_dmat()</code> or <code><a href='https://rdrr.io/r/stats/dist.html'>stats::dist()</a></code></dd>


<dt>clusters</dt>
<dd>output of <code>compute_clusters()</code> or <code>fastcluster::hclust()</code></dd>


<dt>metric_name</dt>
<dd>"silhouette" or "dunn"</dd>


<dt>max_k</dt>
<dd>maximum number of clusters to cut using <code>dendextend::cutree()</code>. Default is 14.</dd>

</dl>

compute_metric: Compute an internal evaluation metric for clustered data

Description

Usage

Value

Arguments

Examples