prior.GP {plgp}

Generate priors for GP models

Description

Generate priors for Gaussian process (GP) regression, classification, or combined unknown constraint models

Usage

prior.GP(m, cov = c("isotropic", "separable", "sim"))
prior.CGP(m, cov = c("isotropic", "separable", "sim"))
prior.ConstGP(m, cov.GP = c("isotropic", "separable", "sim"),
              cov.CGP = cov.GP)

Arguments

m

positive scalar integer specifying the dimensionality of the input space

cov

whether to use an "isotropic" or "separable" power exponential correlation function with power 2 – nugget included; a single index model ("sim") capability is provided as “beta” functionality; applies to both regression and classification GPs

cov.GP

specifies the covariance for the real-valued response in the combined unknown constraint GP model

cov.CGP

specifies the covariance for the categorical response in the combined unknown constraint GP model

Details

These function generate a default prior object in the correct format for use with the other PL routines, e.g., init.GP and pred.GP. The object returned may be modified as necessary.

The prior.ConstGP is essentially the combination of prior.GP and prior.CGP for regression and classification GP models, respectively

Value

a valid prior object for the appropriate GP model;

By making the output $drate and/or $grate values negative causes the corresponding lengthscale d parameter(s) and nugget d parameter to be fixed at the reciprocal of their absolute values, respectively. This effectively turns off inference for these values, and allows one to study the GP predictive distribution as a function of fixed values. When both are fixed it is sensible to use only one particle (P=1, as an argument to PL)

Author(s)

Robert B. Gramacy, [e-mail]

References

Gramacy, R. and Polson, N. (2011). “Particle learning of Gaussian process models for sequential design and optimization.” Journal of Computational and Graphical Statistics, 20(1), pp. 102-118; arXiv:0909.5262

Gramacy, R. and Lee, H. (2010). “Optimization under unknown constraints”. Bayesian Statistics 9, J. M. Bernardo, M. J. Bayarri, J. O. Berger, A. P. Dawid, D. Heckerman, A. F. M. Smith and M. West (Eds.); Oxford University Press

Gramacy, R. (2020). “Surrogates: Gaussian Process Modeling, Design and Optimization for the Applied Sciences”. Chapman Hall/CRC; https://bobby.gramacy.com/surrogates/

https://bobby.gramacy.com/r_packages/plgp/

See Also

PL, lpredprob.GP, propagate.GP, init.GP, pred.GP

Examples

## See the demos via demo(package="plgp") and the examples
## section of ?plgp