fda (version 2.4.0)

register.newfd: Register Functional Data Objects with Pre-Computed Warping Functions

Description

This function registers a new functional data object to pre-computed warping functions.

Usage

register.newfd(yfd, Wfd,type=c('direct','monotone','periodic'))

Arguments

yfd

a multivariate functional data object defining the functions to be registered with Wfd.

Wfd

a functional data object defining the registration functions to be used to register yfd. This can be the result of either landmarkreg or register.fd.

type

indicates the type of registration function.

• direct

{ assumes Wfd is a direct definition of the registration functions. This is produced by landmarkreg. } monotone{ assumes that

Value

• a functional data object defining the registered curves.

source

Ramsay, James O., and Silverman, Bernard W. (2006), Functional Data Analysis, 2nd ed., Springer, New York.

Ramsay, James O., and Silverman, Bernard W. (2002), Applied Functional Data Analysis, Springer, New York, ch. 6 & 7.

Details

Only shift registration is considered for the periodic case.

See Also

landmarkreg, register.fd

Examples

# Register the lip data with landmark registration, then register the first
# derivatives with the pre-computed warping functions.

# Lip data:
lipfd <- smooth.basisPar(liptime, lip, 6, Lfdobj=int2Lfd(4),
                         lambda=1e-12)$fd
names(lipfd$fdnames) <- c("time(seconds)", "replications", "mm")

# Landmark Registration:

lipmeanmarks <- apply(lipmarks,2,mean)

wnbasis <- 6
wnorder <- 4
wbreaks <- c(0,lipmeanmarks,0.35)

warpbasis <- create.bspline.basis(nbasis=wnbasis, norder=wnorder,
                                  breaks=wbreaks);
WfdPar    <- fdPar(fd(basisobj=warpbasis), 2, 1e-4)

lipreglist <- landmarkreg(lipfd, as.matrix(lipmarks), lipmeanmarks, WfdPar)

# And warp:

Dlipfd = deriv.fd(lipfd,Lfdobj=1)

Dlipregfd = register.newfd(Dlipfd,lipreglist$warpfd,type='direct')