fdaMixed-package: Functional Data Analysis in a Mixed Model Framework

In fdaMixed: Functional Data Analysis in a Mixed Model Framework

Functional Data Analysis in a Mixed Model Framework

Description

Likelihood based analysis of 1-dimension functional data in a mixed-effets model framework. The methodology is designed for equidistantly sampled high frequency data, where the needed matrix computation may be approximated by semi-explicit operator equivalents with linear computational complexity. Extensions exist for non-equidistantly sampled data, but these have not been implemented.

Author(s)

Bo Markussen <[e-mail]>

References

Bo Markussen (2013), "Functional data analysis in an operator based mixed model framework", Bernoulli, vol. 19, pp. 1-17.

Conrad Sanderson (2010), "Armadillo: An open source C++ linear algebra library for fast prototyping and computationally intensive experiments", NICTA technical report.

Dirk Eddelbuettel, "Rcpp: Seamless R and C++ Integration with Rcpp", UseR!, Springer, 2013.

See Also

Implementation done using the package RcppArmadillo. For penalized likelihood analysis of functional data see the packages fda and fda.usc.

Examples

x <- seq(0,2*pi,length.out=200)
y.true <- sin(x)+x
y.obs <- y.true + rnorm(200)
est0 <- fdaLm(y.obs~0,Fright="open",right=2*pi)
est1 <- fdaLm(y.obs~0+x,Fright="open",right=2*pi)
plot(x,y.obs,main="Estimating the sum of a line and a curve")
lines(x,y.true,lty=2)
lines(x,est0$xBLUP[,1,1],col=2)
lines(x,est1$betaHat*x+est1$xBLUP[,1,1],col=3)
legend("topleft",c("True curve","Smooth","Line + smooth"),col=1:3,lty=c(2,1,1))

fdaMixed documentation built on Sept. 14, 2023, 1:09 a.m.