EMA (version 1.4.7)

runPCA: Perform an Principal Component Analysis

Description

This function performs a Principal Component Analysis (PCA) and represents the samples or the variables of the analysis.

Usage

runPCA(X, ncp=5, scale=TRUE, ind.sup=NULL, quanti.sup=NULL,quali.sup=NULL,
sample.qual=TRUE, variable.qual=FALSE, sample.cont=TRUE,variable.cont=FALSE,
plotSample=TRUE, plotVariable=FALSE, plotInertia = TRUE, plotBiplot=FALSE,
lab.sample="quality", lab.var=NULL,palette="rainbow",
lim.cos2.sample=0, lim.cos2.var=0, pdf=FALSE, pdfname= NULL, verbose=FALSE, ...)

Arguments

X

a data frame with n rows (samples) and p columns (variables)

ncp

number of dimensions kept in the results (by default 5)

scale

a boolean, if TRUE (value set by default) then data are scaled to unit variance

ind.sup

a vector indicating the indexes of the supplementary individuals

quanti.sup

a vector indicating the indexes of the quantitative supplementary variables

quali.sup

a vector indicating the indexes of the qualitative supplementary variables

sample.qual

a boolean, if TRUE quality sample is displayed, by default = TRUE

variable.qual

a boolean, if TRUE quality variable is displayed, by default = FALSE

sample.cont

a boolean, if TRUE sample contribution is displayed, by default = TRUE

variable.cont

a boolean, if TRUE variable contribution is displayed, by default = FALSE

plotSample

a boolean, if TRUE samples are displayed, by default = TRUE

plotVariable

a boolean, if TRUE variables are displayed, by default = FALSE

plotInertia

a boolean, if TRUE inertia percentage of components is displayed, by default = TRUE

plotBiplot

a boolean, if TRUE biplot is displayed, by default = FALSE

lab.sample

a vector, sample representation is colored by label.sample, by default = NULL

lab.var

a vector, variable representation is colored by label.var, by default = "quality"

palette

character, name of palette color, by default = "rainbow"

lim.cos2.sample

a numeric, for graphics, keep samples with cos2 >= lim.cos2.sample, by default = 0

lim.cos2.var

a numeric, for graphics, keep variables with cos2 >= lim.cos2.var, by default = 0

pdf

a boolean, if TRUE save all the graphics in a pdf file, by default = FALSE

pdfname

pdf file name for saving graphics

verbose

print results if verbose = TRUE, by default = FALSE

...

Arguments to be passed to methods, such as graphical parameters (see 'par').

Value

eig

a matrix containing all the eigenvalues, the percentage of variance and the cumulative percentage of variance

var

a list of matrices containing all the results for the active variables (coordinates, correlation between variables and axes, square cosine, contributions

ind

a list of matrices containing all the results for the active individuals (coordinates, square cosine, contributions)

Returns the individuals factor map for axes 1 and 2, 1 and 3, 2 and 3 Returns the inertia percentage of components By default, print sample coordinates, sample quality and sample contribution

See Also

plotSample, plotVariable, plotInertia

Examples

# NOT RUN {
data("marty")

## PCA on sample with inertia plot and sample plot colored by tumour type

example.subset <- marty[1:100,]
# }
# NOT RUN {
pca <- runPCA(t(example.subset), verbose = FALSE, lab.sample = marty.type.cl)
# }