plotpca: Plotting a Principal Component Analysis

In multifluo: Dealing with Several Images of a Same Object Constituted of Different Zones

Description

Plots the Principal Component Analysis from the pca function, and custumizes the output with

Usage

plotpca(x,map="b",n="a",l=0.05,col=NULL,calc="Chi",epd=NULL,et="bar",dif="n",dax="all")

Arguments

x	results of pca
map	if "bil", biplot is plotted with supplementary individuals projected, and labels of supplementary individuals are written if "bip", biplot is plotted with supplementary individuals projected, and points of supplementary individuals are plotted if "b", biplot is plotted if "til", two-way pca is plotted with supplementary individuals projected, and labels of supplementary individuals are written if "tip", two-way pca is plotted with supplementary individuals projected, and la points of supplementary individuals are plotted if "t", two-way pca is plotted
n	number of axes to plot. By default, "a", which calculates an automatic theoretical number of dimensions (when the contribution of loadings is higher than 1)
l	by default 0.05. Risk to be taken in the agreement ellipses
et	"bar" for the confidence ellipses around the product means "ind" for the representation of a proportion of 1-alpha of the panelist scores
col	select the colors of the different individuals
calc	"Chi", "F" or "Sas". Indicates the type of ellipse to be calculated.
epd	when a biplot is chosen, the arrows can be oversized.
dif	if "n", no tests are run if "h", hotelling test is run to detect differences between zones. if "hc", the p-value is corrected with the bonferroni Correction (l is divided by k(k+1)/ where k is the number of zones) if "s", srivastava test is run to detect differences between zones. if "sc", the p-value is corrected with the bonferroni Correction (l is divided by k(k+1)/ where k is the number of zones) When a test is run, zones not significantly difference are related by a segment
dax	if "all", the test is run on all the Principal Components. If "sig", it is done only on the significant Principal Components. dax can also be a vector determining the index of Principal Component to be chosen for the test (ex dax=c(1,2)).

Examples

data(df.scaled)
resPCA=pca(data=df.scaled[,-1], zone="zone",pixel="pixel")
couleurs=find.colors(rownames(resPCA$IndivCoord)) 
plotpca(x=resPCA)

multifluo documentation built on May 2, 2019, 1:07 p.m.