plot.fssa: Plot Functional Singular Spectrum Analysis Objects
In jtrinka/Rfssa_dev: Functional Singular Spectrum Analysis

Description Usage Arguments Note See Also Examples

This is a plotting method for objects of class functional singular spectrum analysis (fssa). The method is designed to help the user make decisions on how to do the grouping stage of univariate or multivariate functional singular spectrum analysis.

## S3 method for class 'fssa'
plot(
  x,
  d = length(x$values),
  idx = 1:d,
  idy = idx + 1,
  contrib = TRUE,
  groups = as.list(1:d),
  lagvec = 1,
  type = "values",
  var = 1L,
  ylab = NA,
  animation = FALSE,
  ...
)

`x`	an object of class `fssa`
`d`	an integer which is the number of elementary components in the plot
`idx`	a vector of indices of eigen elements to plot
`idy`	a second vector of indices of eigen elements to plot (for type="paired")
`contrib`	a logical where if the value is 'TRUE' (the default), the contribution of the component to the total variance is displayed
`groups`	a list or vector of indices determines grouping used for the decomposition(for type="wcor")
`type`	the type of plot to be displayed where possible types are: `"values"` plot the square-root of singular values (default) `"paired"` plot the pairs of eigenfunction's coefficients (useful for the detection of periodic components) `"wcor"` plot the W-correlation matrix for the reconstructed objects `"vectors"` plot the eigenfunction's coefficients (useful for the detection of period length) `"lcurves"` plot of the eigenfunctions (useful for the detection of period length) `"lheats"` heatmap plot the eigenfunctions (useful for the detection of meaningful patterns) `"periodogram"` periodogram plot (useful for the detecting the frequencies of oscillations in functional data)
`var`	an integer specifying the variable number
`ylab`	the character vector of name of variables
`...`	arguments to be passed to methods, such as graphical parameters

for a multivariate example, see the examples in fssa

fssa, plot.fts

## Not run: 
## Simulated Data Example
require(Rfssa)
require(fda)
n <- 50 # Number of points in each function.
d <- 9
N <- 60
sigma <- 0.5
set.seed(110)
E <- matrix(rnorm(N*d,0,sigma/sqrt(d)),ncol = N, nrow = d)
basis <- create.fourier.basis(c(0, 1), d)
Eps <- fd(E,basis)
om1 <- 1/10
om2 <- 1/4
f0 <- function(tau, t) 2*exp(-tau*t/10)
f1 <- function(tau, t) 0.2*exp(-tau^3) * cos(2 * pi * t * om1)
f2 <- function(tau, t) -0.2*exp(-tau^2) * cos(2 * pi * t * om2)
tau <- seq(0, 1, length = n)
t <- 1:N
f0_mat <- outer(tau, t, FUN = f0)
f0_fd <- smooth.basis(tau, f0_mat, basis)$fd
f1_mat <- outer(tau, t, FUN = f1)
f1_fd <- smooth.basis(tau, f1_mat, basis)$fd
f2_mat <- outer(tau, t, FUN = f2)
f2_fd <- smooth.basis(tau, f2_mat, basis)$fd
Y_fd <- f0_fd+f1_fd+f2_fd
L <-10
U <- fssa(Y_fd,L)
plot(U)
plot(U,d=4,type="lcurves")
plot(U,d=4,type="vectors")
plot(U,d=5,type="paired")
plot(U,d=5,type="wcor")
plot(U,d=5,type="lheats")
plot(U,d=5,type="periodogram")

## End(Not run)