project_pca: Project new data in existing PCA space

In nchlis/pca.utils: R package for easy plotting of PCA results

Description

inputs:

Usage

project_pca(Xnew = NULL, pc = NULL)

Arguments

Xnew:	new data matrix to project.
pc:	prcomp object used for projection generated by prcomp. Must have the same number of columns as X.

Value

Xpc: a matrix that corresponds to the projection of X in the PCA space. Xpc[,1] is the first PC, Xpc[,2] the second PC, etc.

Examples

set.seed(1)
data(iris)

#randomly select 100 data points from the iris data set
#to perform PCA
ix=rep(FALSE,nrow(iris))
ix[sample(nrow(iris),100)]=TRUE

# perform PCA
pc=stats::prcomp(iris[ix,1:4])

# plot the PCA
plot_pca(pc,col=iris$Species[ix])

# project the remaining 50 data points (given by !ix) in the existing PCA space
Xpc=project_pca(iris[!ix,1:4],pc)

#plot the new data points in the existing plot
points(Xpc[,1],Xpc[,2])

nchlis/pca.utils documentation built on May 23, 2019, 1:06 p.m.