cubicbs {blapsr}R Documentation

Construct a cubic B-spline basis.

Description

Computation of a cubic B-spline basis matrix.

Usage

cubicbs(x, lower, upper, K)

Arguments

x

A numeric vector containing the values on which to evaluate the B-spline basis.

lower, upper

The lower and upper bounds of the B-spline basis domain. Must be finite with lower < upper.

K

A positive integer specifying the number of B-spline functions in the basis.

Value

An object of class cubicbs for which print and plot methods are available. The cubicbs class consists of a list with the following components:

x

A numeric vector on which the basis is evaluated.

lower, upper

The lower and upper bounds of the basis domain.

K

The number of cubic B-spline functions in the basis.

knots

The knot sequence to build the basis.

nknots

Total number of knots.

dimbasis

The dimension of the B-spline basis matrix.

Bmatrix

The B-spline basis matrix.

The print method summarizes the B-spline basis and the plot method gives a graphical representation of the basis with dashed vertical lines indicating knot placement and blue ticks the coordinates of x.

Author(s)

Oswaldo Gressani oswaldo_gressani@hotmail.fr.

The core algorithm of the cubicbs function owes much to a code written by Phlilippe Lambert.

References

Eilers, P.H.C. and Marx, B.D. (1996). Flexible smoothing with B-splines and penalties. Statistical Science, 11(2): 89-121.

Examples

lb <- 0  # Lower bound
ub <- 1  # Upper bound
xdom <- runif(100, lb, ub) # Draw uniform values between lb and ub
Bsmat <- cubicbs(xdom, lb, ub, 25) # 100 x 25 B-spline matrix
Bsmat
plot(Bsmat) # Plot the basis

[Package blapsr version 0.6.1 Index]