matrix of sizes \(M \times N\), specifying values of \(N\)
function of \(M\) time points
time
vector of length \(M\), specifying the sample points of functions
lambda
controls amount of warping (default = 0)
pen
alignment penalty (default = "roughness") options are
second derivative ("roughness"), \(L^2\) distance of the warping
function from id ("l2gam"), \(L^2\) distance of the SRVF of the warping
function from that of id ("l2psi"), geodesic distance from id
("geodesic"), and no penalty ("none"). "norm" is kept for backward
compatibility as an alias for "l2gam". The penalty is weighted by
lambda, so it has no effect when lambda = 0.
parallel
run computation in parallel (default = TRUE)
References
T. Harris, J. D. Tucker, B. Li, and L. Shand, "Elastic depths for
detecting shape anomalies in functional data," Technometrics,
10.1080/00401706.2020.1811156, 2020.