Gaussian Kernel R6 class
Gaussian Kernel R6 class
Gaussian Kernel R6 class
R6Class object.
Object of R6Class with methods for fitting GP model.
GauPro::GauPro_kernel -> GauPro_kernel_product
k1kernel 1
k2kernel 2
k1_param_lengthparam length of kernel 1
k2_param_lengthparam length of kernel 2
k1plparam length of kernel 1
k2plparam length of kernel 2
s2Variance
new()
Initialize kernel
kernel_product$new(k1, k2)
k1Kernel 1
k2Kernel 2
k()
Calculate covariance between two points
kernel_product$k(x, y = NULL, params, ...)
xvector.
yvector, optional. If excluded, find correlation of x with itself.
paramsparameters to use instead of beta and s2.
...Not used
param_optim_start()
Starting point for parameters for optimization
kernel_product$param_optim_start(jitter = F, y)
jitterShould there be a jitter?
yOutput
param_optim_start0()
Starting point for parameters for optimization
kernel_product$param_optim_start0(jitter = F, y)
jitterShould there be a jitter?
yOutput
param_optim_lower()
Lower bounds of parameters for optimization
kernel_product$param_optim_lower()
param_optim_upper()
Upper bounds of parameters for optimization
kernel_product$param_optim_upper()
set_params_from_optim()
Set parameters from optimization output
kernel_product$set_params_from_optim(optim_out)
optim_outOutput from optimization
dC_dparams()
Derivative of covariance with respect to parameters
kernel_product$dC_dparams(params = NULL, C, X, C_nonug, nug)
paramsKernel parameters
CCovariance with nugget
Xmatrix of points in rows
C_nonugCovariance without nugget added to diagonal
nugValue of nugget
C_dC_dparams()
Calculate covariance matrix and its derivative with respect to parameters
kernel_product$C_dC_dparams(params = NULL, X, nug)
paramsKernel parameters
Xmatrix of points in rows
nugValue of nugget
dC_dx()
Derivative of covariance with respect to X
kernel_product$dC_dx(XX, X)
XXmatrix of points
Xmatrix of points to take derivative with respect to
s2_from_params()
Get s2 from params vector
kernel_product$s2_from_params(params, s2_est = self$s2_est)
paramsparameter vector
s2_estIs s2 being estimated?
clone()
The objects of this class are cloneable with this method.
kernel_product$clone(deep = FALSE)
deepWhether to make a deep clone.
k1 <- Exponential$new(beta=1) k2 <- Matern32$new(beta=2) k <- k1 + k2 k$k(matrix(c(2,1), ncol=1))
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.