RobustGaSP (version 0.6.6)

Sample: Sample for Robust GaSP model

Description

Function to sample Robust GaSP after the Robust GaSP model has been constructed.

Usage

# S4 method for rgasp
Sample(object, testing_input, num_sample=1,
testing_trend= matrix(1,dim(testing_input)[1],1),
r0=NA, rr0=NA, sample_data=T,
...)

Value

The returned value is a matrix where each column is a sample on the prespecified inputs.

Arguments

object

an object of class rgasp.

testing_input

a matrix containing the inputs where the rgasp is to sample.

num_sample

number of samples one wants.

testing_trend

a matrix of mean/trend for prediction.

r0

the distance between input and testing input. If the value is NA, it will be calculated later. It can also be specified by the user. If specified by user, it is either a matrix or list. The default value is NA.

rr0

the distance between testing input and testing input. If the value is NA, it will be calculated later. It can also be specified by the user. If specified by user, it is either a matrix or list. The default value is NA.

...

Extra arguments to be passed to the function (not implemented yet).

Author

tools:::Rd_package_author("RobustGaSP")

Maintainer: tools:::Rd_package_maintainer("RobustGaSP")

References

Mengyang Gu. (2016). Robust Uncertainty Quantification and Scalable Computation for Computer Models with Massive Output. Ph.D. thesis. Duke University.

Examples

  #------------------------
  # a 1 dimensional example
  #------------------------
  
###########1dim higdon.1.data 
p1 = 1     ###dimensional of the inputs
dim_inputs1 <- p1
n1 = 15   ###sample size or number of training computer runs you have 
num_obs1 <- n1
input1 = 10*matrix(runif(num_obs1*dim_inputs1), num_obs1,dim_inputs1) ##uniform
#####lhs is better
#library(lhs)
#input1 = 10*maximinLHS(n=num_obs1, k=dim_inputs1)  ##maximin lhd sample
output1 = matrix(0,num_obs1,1)
for(i in 1:num_obs1){
  output1[i]=higdon.1.data (input1[i])
}





m1<- rgasp(design = input1, response = output1, lower_bound=FALSE)

#####locations to samples
testing_input1 = seq(0,10,1/50) 
testing_input1=as.matrix(testing_input1)
#####draw 10 samples
m1_sample=Sample(m1,testing_input1,num_sample=10)

#####plot these samples
matplot(testing_input1,m1_sample, type='l',xlab='input',ylab='output')
lines(input1,output1,type='p')