Calculates the variance matrix of the random effects for a natural cubic smoothing spline. It is the tri-diagonal matrix \(\bold{G}_s\) given by Verbyla et al., (1999) multiplied by the variance component for the random spline effects.
mat.ncssvar(sigma2s = 1, knot.points, print = FALSE)
A matrix
containing the variances and covariances of the
random spline effects.
A numeric
giving the value of the variance component
for the random spline effects. The smoothing parameter is then the inverse
of the ratio of this component to the residual variance.
A numeric
giving the values of the knots point
used in fitting the spline. These must be orderd in increasing order.
A logical
indicating whether to print the matrix.
Chris Brien
Verbyla, A. P., Cullis, B. R., Kenward, M. G., and Welham, S. J. (1999). The analysis of designed experiments and longitudinal data by using smoothing splines (with discussion). Journal of the Royal Statistical Society, Series C (Applied Statistics), 48, 269-311.
Zncsspline
.
Gs <- mat.ncssvar(knot.points = 1:10)
Run the code above in your browser using DataLab