mat2fd: Create an 'fd' object from a matrix

In GPFDA: Gaussian Process for Functional Data Analysis

Create an 'fd' object from a matrix

Description

Easy setting up for creating an 'fd' object

Usage

mat2fd(mat, fdList = NULL)

Arguments

mat

Input data, should be a matrix with ncol time points and nrow replications or samples.

fdList

A list with following items: time Sequence of time points (default to be 100 points from 0 to 1). nbasis Number of basis functions used in smoothing, default to be less or equal to 23. norder Order of the functional curves default to be 6. bSpline Logical, if TRUE (default), b-Spline basis is used; otherwise, Fourier basis is used. Pen Default to be c(0,0), meaning that the penalty is on the second order derivative of the curve, since the weight for zero-th and first order derivatives of the curve are set to zero. lambda Smoothing parameter for the penalty. Default to be 1e-4.

Details

All items listed above have default values. If any item is required to change, add that item into the list; otherwise, leave it as NULL. For example, if one only wants to change the number of basis functions, do:

mat2fd(SomeMatrix,list(nbasis=21))

Value

An 'fd' object

References

Ramsay, J., and Silverman, B. W. (2006),

Examples

require(fda)
require(fda.usc)
nrep <- 20   # number of replications
n <- 100     # number of time points
input <- seq(-1, pi, length.out=n) # time points
ry <- rnorm(nrep, sd=10)
y <- matrix(NA, ncol=n, nrow=nrep)
for(i in 1:nrep)  y[i,] <- sin(2*input)*ry[i]

plot.fdata(fdata(y,input))

yfd <- mat2fd(y, list(lambda=0.01))
plot(yfd)

yfd <- mat2fd(y, list(lambda=0.00001))
plot(yfd)

GPFDA documentation built on Sept. 11, 2023, 1:08 a.m.