Global Sensitivity Analysis

This function performs a binning based estimation of the global sensitivity of a model's output with respect to the model's parameters. The output can be a prediction of the model's behaviour in a scenario of interest (parameters, input, intial values, boundary conditions, scheduled events etc.). The output models a potentially measurable value (the "observable"). The sample-rows and the output rows must correspond (they must be from the same model simulation).

Usage

gsa_binning(parSample, outputSample, nBins = "Sturges")

Arguments

parSample

a matrix of parameter vectors (rows)

outputSample

a matrix, with rows of outputs (row-index is the sample index)

nBins

number of bins, if unset defaults to the default of the hist function

Value

sensitivity S[i,j] of output[i] with respect to parameter[j]

Examples

  rprior <- rNormalPrior(c(-1,0,1),c(1,2,3))
  X <- rprior(10000)
  colnames(X) <- LETTERS[seq(3)]
  Z <- exp(X[,1,drop=FALSE]+X[,2,drop=FALSE])
  colnames(Z) <- "alpha"
  GSA <- gsa_binning(X,Z)
  print(GSA)
#>               A         B            C
#> alpha 0.0430478 0.4522683 0.0002483376
  cat("global sensitivity of alpha with respect to B: ",GSA['alpha','B'],"\n")
#> global sensitivity of alpha with respect to B:  0.4522683