R/Kernel.R

In mathieucarmassi/calibrationcode: Code Calibration in a Bayesian Framework

#' A Reference Class to generates differents model objects
#'
#' @description See the function \code{\link{model}} which produces an instance of this class
#' @field code a function which takes in entry X and theta
#' @field X the matrix of the forced variables
#' @field var the variance for the covariance function
#' @field psi the scale vector of correlation length
#' @field n number of experiments
#' @field Kernel.type the chosen form of covariance
#' @field Cov the covariance
#' @export
Kernel.class <- R6Class(classname = "Kernel.class",
                           public = list(
                             X           = NULL,
                             var         = NULL,
                             psi         = NULL,
                             Kernel.type = NULL,
                             Cov         = NULL,
                             initialize = function(X=NA,var=NA,psi=NA,Kernel.type=NA)
                             {
                               self$X           <- X
                               self$var         <- var
                               self$psi         <- psi
                               self$Kernel.type <- Kernel.type
                             }
                           ))


gauss.class <- R6Class(classname= "gauss.class",
                           inherit = Kernel.class,
                           public = list(
                             initialize = function(X,var,psi,Kernel.type="gauss")
                             {
                               super$initialize(X,var,psi,Kernel.type="gauss")
                               self$Cov <- self$covariance(X,var,psi)
                             },
                             covariance = function(X,var,psi)
                             {
                               m <- as.matrix(dist(X))
                               self$Cov <- var*exp(-1/2*(m/psi)^2)
                               return(self$Cov)
                             }
                           ))

exp.class <- R6Class(classname= "exp.class",
                           inherit = Kernel.class,
                         public = list(
                           initialize = function(X,var,psi,Kernel.type="exp")
                           {
                             super$initialize(X,var,psi,Kernel.type="exp")
                             self$Cov <- self$covariance(X,var,psi)
                           },
                           covariance = function(X,var,psi)
                           {
                             m <- as.matrix(dist(X))
                             self$Cov <- var*exp(m/psi)
                             return(self$Cov)
                           }
))


matern3_2.class <- R6Class(classname= "matern3_2.class",
                         inherit = Kernel.class,
                         public = list(
                         initialize = function(X,var,psi,Kernel.type="matern3_2")
                         {
                           super$initialize(X,var,psi,Kernel.type="matern3_2")
                           self$Cov <- self$covariance(X,var,psi)
                         },
                         covariance = function(X,var,psi)
                         {
                           m <- as.matrix(dist(X))
                           self$Cov <- var*(1+sqrt(3)*m/psi)*exp(-sqrt(3)*m/psi)
                           return(self$Cov)
                         }
))


matern5_2.class <- R6Class(classname= "matern5_2.class",
                               inherit = Kernel.class,
                               public = list(
                               initialize = function(X,var,psi,Kernel.type="matern5_2")
                               {
                                 super$initialize(X,var,psi,Kernel.type="matern5_2")
                                 self$Cov <- self$covariance(X,var,psi)
                               },
                               covariance = function(X,var,psi)
                               {
                                 m <- as.matrix(dist(X))
                                 self$Cov <- var*(1+sqrt(5)*m/psi+5/3*(m/psi)^2)*exp(-sqrt(5)*m/psi)
                                 return(self$Cov)
                               }
))