getUniCl: Unique cluster labels

In funcy: Functional Clustering Algorithms

Description

For repeated measurements on the same ID, build unique cluster labels according to the vector of IDs.

Usage

getUniCl(id, clusters, reduce=TRUE)

Arguments

id	Vector of IDs.
clusters	Vector of cluster labels.
reduce	TRUE if cluster labels shall be reduced according to unique IDs. FALSE if cluster labels shall be duplicated according to IDs.

Details

Data might have been stored in 4 columns: curveID, curve evaluations, time points and cluster labels. Cluster labels were therefore repeated for each curve evaluation point. Method funcit accepts dataset only in formats "Format1" and "Format2" and an optional vector of cluster true labels clusters of length nr_curves. getUniCl can be applied to columns curveID and repeated labels to reduce them to the number of curves.

Value

A vector of either reduced (reduce=TRUE) or duplicated (reduce=FALSE) cluster labels.

Author(s)

Christina Yassouridis

Examples

##Generate dataset
nr_time <- sample(1:5, 100,  replace=TRUE)
clusters <- sample(1:4, 100, replace=TRUE)
IDs <- rep(1:100, nr_time)
cls <- rep(clusters,nr_time)
IDs
cls

##Get reduced cluster vector according to IDs
unicl <- getUniCl(IDs,cls)
unicl

##Get original cluster vector from reduced one
dupcl <- getUniCl(IDs,unicl, reduce=FALSE)
dupcl

funcy documentation built on May 2, 2019, 9:22 a.m.