在 R 中组合或合并两个或多个 .fd 类功能数据

Question

I am a newcomer in functional data analysis (FDA).

library(fda)

set.seed(151)

I1 <- matrix(rnorm(20*20,mean=0,sd=1),20,20)
I2 <- matrix(rnorm(15*20,mean=0.5,sd=1),15,25)

data1 <- t(I1)
data2 <- t(I2)
minutetime1 <- seq(from = 1, to = 25, length.out = 20)
minutetime2 <- seq(from = 1, to = 25, length.out = 25)

minutebasis <- create.bspline.basis(rangeval=c(0,25),nbasis=10)

fd1<- Data2fd(data1, minutetime1, basisobj=minutebasis)
is.fd(fd1)
# [1] TRUE
fd2<- Data2fd(data2, minutetime2, basisobj=minutebasis)
is.fd(fd2)
# [1] TRUE

I would like to merge or combine fd1 and fd2 ( like combining two vectors), and the result will also be the .fd class. I have used c(fd1,fd2) , merge(fd1,fd2) and modifyList(fd1,fd2) , etc.

Answer 1

Because your data are represented on the same basis, you should be able to create a combined fd object by combining the basis function coefficients from fd1 and fd2 .

The coefficients are stored in matrices. Rows correspond to basis functions and columns correspond to observations (see ?fd ), so we can bind columns to combine observations.

merged_coefs <- cbind(fd1$coefs, fd2$coefs)
merged_fd <- fd(coef = merged_coefs, basisobj = minutebasis)
is.fd(merged_fd)