我目前正在尝试绘制第一和第二谐波（如Ramsay - FDA中使用R和MATLAB所描述的）。理想情况下，它应该看起来像下面的图像-x1c 0d1x到目前为止，这是我的设置-

handpartx <- handwrit[,,1]
handparty <- handwrit[,,2]
curv.Lfd <- int2Lfd(2)
lambda <- 1e-2
breaks <- seq(min(handwritTime), max(handwritTime), length.out = 100)
xbasis <- create.bspline.basis(range(handwritTime), norder = 6,
                               breaks = breaks)
ybasis <- create.bspline.basis(range(handwritTime), norder = 6,
                               breaks = breaks)

curvx.fdPar <- fdPar(xbasis, curv.Lfd, lambda)
curvy.fdPar <- fdPar(ybasis, curv.Lfd, lambda)

handxsmooth <- smooth.basis(handwritTime, handpartx, curvx.fdPar)
handysmooth <- smooth.basis(handwritTime, handparty, curvy.fdPar)
mean_x <- eval.fd(fdobj = mean.fd(handxsmooth$fd), evalarg = handwritTime)
mean_y <- eval.fd(fdobj = mean.fd(handysmooth$fd), evalarg = handwritTime)

字符串
对于平均FDA脚本和联合PCA -

zbasis <- smooth.basis(handwritTime, handwrit, curvx.fdPar)$fd
zbasis$fdnames[[1]] <- "Time"
zbasis$fdnames[[2]] <- "Replications"
zbasis$fdnames[[3]] <- list("X", "Y")

nharm <- 10
handPCA <- pca.fd(zbasis, nharm)

型
有人能指出我是如何把谐波的变化加到平均值上的吗？或者更好的方法来重写我现在的代码。提前感谢！