q1Symm: Qualitative Correlation or Covariance Kernel with one Input...

In kergp: Gaussian Process Laboratory

Description

Qualitative correlation or covariance kernel with one input and general symmetric correlation.

Usage

  q1Symm(factor, input = "x", cov = c("corr", "homo", "hete"), intAsChar = TRUE)

Arguments

factor	A factor with the wanted levels for the covariance kernel object.
input	Name of (qualitative) input for the kernel.
cov	Character telling if the result is a correlation kernel, an homoscedastic covariance kernel or an heteroscedastic covariance kernel with an arbitrary variance vector.
intAsChar	Logical. If TRUE (default), an integer-valued input will be coerced into a character. Otherwise, it will be coerced into a factor.

Value

An object with class "covQual" with d = 1 qualitative input.

Note

Correlation kernels are needed in tensor products because the tensor product of two covariance kernels each with unknown variance would not be identifiable.

See Also

The corLevSymm function used to compute the correlation matrix and its gradients w.r.t. the correlation parameters.

Examples

School <- factor(1L:3L, labels = c("Bad", "Mean" , "Good"))
myCor <- q1Symm(School, input = "School")
coef(myCor) <- c(theta_2_1 = pi / 3, theta_3_1 = pi / 4, theta_3_2 = pi / 8)
plot(myCor, type = "cor")

## Use a data.frame with a factor
set.seed(246)
newSchool <- factor(sample(1L:3L, size = 20, replace = TRUE),
                    labels = c("Bad", "Mean" , "Good"))
C1 <- covMat(myCor, X = data.frame(School = newSchool),
             compGrad = FALSE, lowerSQRT = FALSE)

kergp documentation built on March 18, 2021, 5:06 p.m.