plot.pca.fd {fda} | R Documentation |
Display the types of variation across a sample of functions. Label with the eigenvalues that indicate the relative importance of each mode of variation.
plot.pca.fd(x, nx = 128, pointplot = TRUE, harm = 0, expand = 0, cycle = FALSE, ...)
x |
a functional data object. |
nx |
Number of points to plot or vector (if length > 1) to use as
evalarg in evaluating and plotting the functional principal
components.
|
pointplot |
logical: If TRUE, the harmonics / principal components are plotted as '+' and '-'. Otherwise lines are used. |
harm |
Harmonics / principal components to plot. If 0, plot all.
If length(harm) > sum(par("mfrow")), the user advised, "Waiting to confirm page change..." and / or 'Click or hit ENTER for next page' for each page after the first. |
expand |
nonnegative real: If expand == 0 then effect of +/- 2 standard deviations of each pc are given otherwise the factor expand is used. |
cycle |
logical: If cycle=TRUE and there are 2 variables then a cycle plot will be drawn If the number of variables is anything else, cycle will be ignored. |
... |
other arguments for 'plot'. |
Produces one plot for each principal component / harmonic to be plotted.
invisible(NULL)
# carry out a PCA of temperature # penalize harmonic acceleration, use varimax rotation daybasis65 <- create.fourier.basis(c(0, 365), nbasis=65, period=365) harmaccelLfd <- vec2Lfd(c(0,(2*pi/365)^2,0), c(0, 365)) harmfdPar <- fdPar(daybasis65, harmaccelLfd, lambda=1e5) daytempfd <- data2fd(CanadianWeather$dailyAv[,,"Temperature.C"], day.5, daybasis65, argnames=list("Day", "Station", "Deg C")) daytemppcaobj <- pca.fd(daytempfd, nharm=4, harmfdPar) # plot harmonics, asking before each new page after the first: plot.pca.fd(daytemppcaobj) # plot 4 on 1 page op <- par(mfrow=c(2,2)) plot.pca.fd(daytemppcaobj, cex.main=0.9) par(op)