selection_p: Optimal Sampling Design for Functional Data Analysis -...

In soyoung-park/FDAdesign: Optimal Sampling Design for Functional Data

Description

Selects number of optimal points, p, from a vector of p's (p_vec) given as input. This function also returns results of the opt_design_fda function for given p's (p_vec).

Usage

selection_p(p_vec = 1:5, threshold = 5, Phi, lambda, B = NULL,
  sigma2 = 10^-14)

Arguments

p_vec	a vector of p's
threshold	user-specified threshold h for selecting p (relative error levels)
Phi	d by L matrix of eigenfunctions evaluated at d candidate points; L is number of PCs.
lambda	eigenvalues; a vector of length L
B	design criterion matrix (e.g. for recovering curves, B = diag(L); a square matrix with dim = L
sigma2	measurement error variance associated with functional object.

Value

plot plot of error levels corresponding to p_vec.

p.sel number of optimal points, p, determined with given threshold.

opt.sel result of opt_design_fda corresponding to the selected p, p.sel.

opt_result results of opt_design_fda corresponding to p_vec.

INPUT input of selection_p provided as input.

Author(s)

So Young Park [e-mail], Luo Xiao [e-mail], Ana-Maria Staicu [e-mail]

See Also

opt_design_fda / selection_p / interactive_plot

Examples

# see example for opt_design_fda function

soyoung-park/FDAdesign documentation built on May 30, 2019, 6:34 a.m.