qrnn (version 2.0.1)

qrnn.rbf: Radial basis function kernel

Description

Evaluate a kernel matrix based on the radial basis function kernel. Can be used in conjunction with qrnn.fit with Th set to linear and penalty set to a nonzero value for kernel quantile ridge regression.

Usage

qrnn.rbf(x, x.basis, sigma)

Arguments

x

covariate matrix with number of rows equal to the number of samples and number of columns equal to the number of variables.

x.basis

covariate matrix with number of rows equal to the number of basis functions and number of columns equal to the number of variables.

sigma

kernel width

Value

kernel matrix with number of rows equal to the number of samples and number of columns equal to the number of basis functions.

References

Cannon, A.J., 2011. Quantile regression neural networks: implementation in R and application to precipitation downscaling. Computers & Geosciences, 37: 1277-1284. doi:10.1016/j.cageo.2010.07.005

See Also

qrnn.fit

Examples

# NOT RUN {
x <- as.matrix(iris[,"Petal.Length",drop=FALSE])
y <- as.matrix(iris[,"Petal.Width",drop=FALSE])

cases <- order(x)
x <- x[cases,,drop=FALSE]
y <- y[cases,,drop=FALSE]

set.seed(1)
kern <- qrnn.rbf(x, x.basis=x, sigma=1)

parms <- qrnn.fit(x=kern, y=y, tau=0.5, penalty=0.1,
                  Th=linear, Th.prime=linear.prime,
                  iter.max=500, n.trials=1)
p <- qrnn.predict(x=kern, parms=parms)

matplot(x, cbind(y, p), type=c("p", "l"), pch=1, lwd=1)
# }