PCA.DB: PCA.DB

In majpark21/TSexploreR: Explorative Tools for Time Series

Description

Compute the Davies-Bouldin index for a pca clustering. This is used to quantify how much groups are taken away in PCA space.

Usage

PCA.DB(pca, PC = NULL, PCvar = NULL, labels.pca, label.cluster)

Arguments

pca	A pca object.
PC	Which PCs should be used to compute the DB index. Consider only a few components to use only leading trends in data.
PCvar	If provided, replace PC. Instead of using a given set of PC, will use as many as necessary to reach PCvar percentage of explained variance.
labels.pca	A data.table or character matrix with a number of rows equal to the number of time series used in pca. Each column contains a different label for the associated trajectory. Note that column names must
label.cluster	be provided.

Value

A numeric, the Davies-Bouldin index of the clustering.

majpark21/TSexploreR documentation built on Oct. 16, 2019, 2:46 p.m.