rk: kernel function

In fpcb: Predictive Confidence Bands for Functional Time Series Forecasting

Description

Computes the Gram matrix of the gaussian kernel over a grid of values and computes its singular value decomposition.

Usage

rk(grid, sigma = 1, r, tol = 1e-08)

Arguments

grid	grid of points where the kernel function is evaluated.
sigma	is the temperature of the kernel (standard deviation)
r	the dimension of the basis system of the Gran matrix (K). If missing then r is the rank of K.
tol	A tolerance to keep the first d eigenvalues of A. Default = 1e-08.

Value

grid	grid of points where the kernel function is evaluated.
K	Kernel Gram matrix
U	first r eigenvectors of K using svd.
D	first r eigenvectors of K using svd.

Author(s)

J. Cugliari and N. Hernández

Examples

grid = seq(0,1,,100)
rk(grid, sigma = 1)

fpcb documentation built on June 7, 2021, 9:08 a.m.