remoteCalcChol: Calculate Distributed Cholesky Decomposition

Description

remoteCalcChol calculates a distributed Cholesky decomposition from a distributed positive definite matrix. The Cholesky factor and the original matrix can both be contained within environments and ReferenceClass objects as well as the global environment on the slave processes.

Usage

remoteCalcChol(matName, cholName, matPos = '.GlobalEnv', cholPos = '.GlobalEnv', n, h = 1)

Arguments

Details

Computes the distributed Cholesky decomposition using a blocked algorithm similar to that in ScaLapack. When $h$ is 1, the number of blocks, representing the lower triangle of the original matrix and of the Cholesky factor, is equal to the number of processes. For larger values of $h$, there are multiple blocks assigned to each process.

References

Paciorek, C.J., B. Lipshitz, W. Zhuo, Prabhat, C.G. Kaufman, and R.C. Thomas. 2013. Parallelizing Gaussian Process Calculations in R. arXiv:1305.4886. http://arxiv.org/abs/1305.4886.

Examples

Run this code

if(require(fields)) {
  SN2011fe <- SN2011fe_subset
  SN2011fe_newdata <- SN2011fe_newdata_subset
  SN2011fe_mle <- SN2011fe_mle_subset
  nProc <- 3
n <- nrow(SN2011fe)
m <- nrow(SN2011fe_newdata)
nu <- 2
inputs <- c(as.list(SN2011fe), as.list(SN2011fe_newdata), nu = nu)
prob <- krigeProblem$new("prob", numProcesses = nProc, n = n, m = m,
predMeanFunction = SN2011fe_predmeanfunc, crossCovFunction = SN2011fe_crosscovfunc,
predCovFunction = SN2011fe_predcovfunc, meanFunction = SN2011fe_meanfunc,
 covFunction = SN2011fe_covfunc,  inputs = inputs, params = SN2011fe_mle$par,
data = SN2011fe$flux, packages = c("fields"))
remoteCalcChol(matName = 'C', cholName = 'L', matPos = 'prob',
  cholPos = 'prob', n = n, h = prob$h_n)
L <- collectTriangularMatrix('L', objPos = 'prob', n = n, h = prob$h_n)
}