bisquare_fns: Bisquare function

In andrewzm/bicon: Bivariate modelling using the conditional approach

Description

The bisquare function

Usage

bisquare_1d(h, delta = 0, r = 1, A = 1)

bisquare_2d(h1, h2, delta = c(0, 0), r = 1, A = 1)

bisquare_B(h1, h2, delta = c(0, 0), r = 1, A = 1, areas = rep(1,
  length(h1)), n1 = 10L, n2 = 10L)

Arguments

h	displacement (1d)
delta	shift parameter(s)
r	aperture parameter
A	gain parameter (amplitude)
h1	first component of displacement vector (2d)
h2	second component of displacement vector (2d)
areas	area associated with each column in B
n1	number of rows
n2	number of columns

Details

The bisquare function (shifted by Δ) is given by

b(s,v) = A{1 - (|v- s - d|/r)^2}^2 if | v -s - d| <= r, 0 otherwise

The function bisquare_1d accepts h in any shape or size, while for bisquare_2d, h1 and h2 need to be vectors of the same size. The parameter delta needs to be equal to one or two in length, depending on the spatial dimension.

The function bisquare_B is used to construct the matrix B given the bisquare parameters. It is meant to be used in problems of 2 dimensions.

Examples

h <- seq(-10,10,0.1)
y <- bisquare_1d(h=h,delta=3,r=4)
plot(h,y)

andrewzm/bicon documentation built on May 10, 2019, 11:15 a.m.