Emulation of computer code output
This package allows one to estimate the output of a computer program,
as a function of the input parameters, without actually running it.
The computer program is assumed to be a Gaussian process, whose
parameters are estimated using Bayesian techniqes that give a PDF of
expected program output. This PDF is conditional on a “training
set” of runs, each consisting of a point in parameter space and the
model output at that point. The emphasis is on complex codes that take
weeks or months to run, and that have a large number of undetermined
input parameters; many climate prediction models fall into this class.
The emulator essentially determines Bayesian a-postiori estimates of
the PDF of the output of a model, conditioned on results from previous
runs and a user-specified prior linear model. A working example is
given in the help page for function
interpolant(), which should
be the users's first point of reference.
The primary function of the package is
Robin K. S. Hankin
J. Oakley 1999. Bayesian uncertainty analysis for complex computer codes, PhD thesis, University of Sheffield
J. Oakley and A. O'Hagan, 2002. Bayesian Inference for the Uncertainty Distribution of Computer Model Outputs, Biometrika 89(4), pp769-784
R. K. S. Hankin 2005. Introducing BACCO, an R bundle for Bayesian analysis of computer code output, Journal of Statistical Software, 14(16)
1 2 3 4 5 6 7 8 9 10 11 12
# The following example takes a toy dataframe (toy), which represents an # experimental design. Variable d contains observations at points in a # six dimensional parameter space specified by the rows of toy. # Function interpolant() is then called to estimate what the # observation would be at a point that has no direct observation. data(toy) d <- c(11.05, 7.48, 12.94, 14.91, 11.34, 5.0, 11.83, 11.761, 11.62, 6.70) fish <- rep(1,6) x <- rep(0.5, 6) interpolant(x, d, toy, scales=fish,give=TRUE)
Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.