pei {dgpsi}R Documentation

Locate the next design point for a (D)GP emulator or a bundle of (D)GP emulators using PEI

Description

This function searches from a candidate set to locate the next design point(s) to be added to a (D)GP emulator or a bundle of (D)GP emulators using the Pseudo Expected Improvement (PEI), see the reference below.

Usage

pei(object, x_cand, ...)

## S3 method for class 'gp'
pei(object, x_cand, pseudo_points = NULL, batch_size = 1, ...)

## S3 method for class 'dgp'
pei(
  object,
  x_cand,
  pseudo_points = NULL,
  batch_size = 1,
  workers = 1,
  threading = FALSE,
  aggregate = NULL,
  ...
)

## S3 method for class 'bundle'
pei(
  object,
  x_cand,
  pseudo_points = NULL,
  batch_size = 1,
  workers = 1,
  threading = FALSE,
  aggregate = NULL,
  ...
)

Arguments

object

can be one of the following:

  • the S3 class gp.

  • the S3 class dgp.

  • the S3 class bundle.

x_cand

a matrix (with each row being a design point and column being an input dimension) that gives a candidate set from which the next design point(s) are determined. If object is an instance of the bundle class, x_cand could also be a list with the length equal to the number of emulators contained in the object. Each slot in x_cand is a matrix that gives a candidate set for each emulator included in the bundle. See Note section below for further information.

...

any arguments (with names different from those of arguments used in pei()) that are used by aggregate or gp() (for emulating the ES-LOO errors) can be passed here.

pseudo_points

an optional matrix (with columns being input dimensions) that gives the pseudo input points for PEI calculations. See the reference below for further details about the pseudo points. When object is an instance of the bundle class, pseudo_points can also be a list with the length equal to the number of emulators in the bundle. Each element in the list is a matrix that gives the the pseudo input points for the corresponding emulator in the bundle. Defaults to NULL. When pei() is used in design(), pseudo_points will be automatically generated by design().

batch_size

an integer that gives the number of design points to be chosen. Defaults to 1.

workers

the number of workers/cores to be used for the criterion calculation. If set to NULL, the number of workers is set to ⁠(max physical cores available - 1)⁠. Defaults to 1.

threading

a bool indicating whether to use the multi-threading to accelerate the criterion calculation for a DGP emulator. Turning this option on could improve the speed of criterion calculations when the DGP emulator is built with a moderately large number of training data points and the Matérn-2.5 kernel.

aggregate

an R function that aggregates scores of the PEI across different output dimensions (if object is an instance of the dgp class) or across different emulators (if object is an instance of the bundle class). The function should be specified in the following basic form:

  • the first argument is a matrix representing scores. The rows of the matrix correspond to different design points. The number of columns of the matrix equals to:

    • the emulator output dimension if object is an instance of the dgp class; or

    • the number of emulators contained in object if object is an instance of the bundle class.

  • the output should be a vector that gives aggregations of scores at different design points.

Set to NULL to disable the aggregation. Defaults to NULL.

Details

See further examples and tutorials at https://mingdeyu.github.io/dgpsi-R/.

Value

Note

References

Mohammadi, H., Challenor, P., Williamson, D., & Goodfellow, M. (2022). Cross-validation-based adaptive sampling for Gaussian process models. SIAM/ASA Journal on Uncertainty Quantification, 10(1), 294-316.

Examples

## Not run: 

# load packages and the Python env
library(lhs)
library(dgpsi)

# construct a 1D non-stationary function
f <- function(x) {
 sin(30*((2*x-1)/2-0.4)^5)*cos(20*((2*x-1)/2-0.4))
}

# generate the initial design
X <- maximinLHS(10,1)
Y <- f(X)

# training a 2-layered DGP emulator with the global connection off
m <- dgp(X, Y, connect = F)

# generate a candidate set
x_cand <- maximinLHS(200,1)

# locate the next design point using PEI
next_point <- pei(m, x_cand = x_cand)
X_new <- x_cand[next_point,,drop = F]

# obtain the corresponding output at the located design point
Y_new <- f(X_new)

# combine the new input-output pair to the existing data
X <- rbind(X, X_new)
Y <- rbind(Y, Y_new)

# update the DGP emulator with the new input and output data and refit with 500 training iterations
m <- update(m, X, Y, refit = TRUE, N = 500)

# plot the LOO validation
plot(m)

## End(Not run)

[Package dgpsi version 2.4.0 Index]