PLIquantile: Perturbed-Law based sensitivity Indices (PLI) for quantile

Description

PLIquantile computes the Perturbed-Law based Indices (PLI) for quantile, which are robustness indices related to a quantile of a model output, estimated by a Monte Carlo method, See Sueur et al. (2017) and Iooss et al. (2020).

Usage

PLIquantile(order,x,y,deltasvector,InputDistributions,type="MOY",samedelta=TRUE,
            percentage=TRUE,nboot=0,conf=0.95,bootsample=TRUE)

Arguments

order

the order of the quantile to estimate.

the matrix of simulation points coordinates, one column per variable.

the vector of model outputs.

deltasvector

a vector containing the values of delta for which the indices will be computed.

InputDistributions

a list of list. Each list contains, as a list, the name of the distribution to be used and the parameters. Implemented cases so far:

For a mean perturbation: Gaussian, Uniform, Triangle, Left Trucated Gaussian, Left Truncated Gumbel. Using Gumbel requires the package evd.
For a variance perturbation: Gaussian, Uniform.

type

a character string in which the user will specify the type of perturbation wanted. The sense of "deltasvector" varies according to the type of perturbation:

type can take the value "MOY",in which case deltasvector is a vector of perturbated means.
type can take the value "VAR",in which case deltasvector is a vector of perturbated variances, therefore needs to be positive integers.

samedelta

a boolean used with the value "MOY" for type.

If it is set at TRUE, the mean perturbation will be the same for all the variables.
If not, the mean perturbation will be new_mean = mean+sigma*delta where mean, sigma are parameters defined in InputDistributions and delta is a value of deltasvector.

percentage

a boolean that defines the formula used for the PLI.

If it is set at FALSE, the initially proposed formula is used (see Sueur et al., 2017).
If not (set as TRUE), the PLI is given in percentage of variation of the quantile (see Iooss et al., 2020).

nboot

the number of bootstrap replicates.

conf

the confidence level for bootstrap confidence intervals.

bootsample

If TRUE, the uncertainty about the original quantile estimation is taken into account in the PLI confidence intervals (see Iooss et al., 2020). If FALSE, standard confidence intervals are computed for the PLI. It mainly changes the CI at small delta values.

Value

PLIquantile returns a list of matrix (each column corresponds to an input, each line corresponds to a twist of amplitude delta) containing the following components:

PLI

the PLI.

PLICIinf

the bootstrap lower confidence interval values of the PLI.

PLICIsup

the bootstrap upper confidence interval values of the PLI.

quantile

the perturbed quantile.

quantileCIinf

the bootstrap lower confidence interval values of the perturbed quantile.

quantileCIsup

the bootstrap upper confidence interval values of the perturbed quantile.

References

T. Delage, R. Sueur and B. Iooss, 2018, Robustness analysis of epistemic uncertainties propagation studies in LOCA assessment thermal-hydraulic model, ANS Best Estimate Plus Uncertainty International Conference (BEPU 2018), Lucca, Italy, May 13-19, 2018.

C. Gauchy, J. Stenger, R. Sueur and B. Iooss, An information geometry approach for robustness analysis in uncertainty quantification of computer codes, Preprint, https://hal.archives-ouvertes.fr/hal-02425477

B. Iooss, V. Verges and V. Larget, BEPU robustness analysis via perturbed-law based sensitivity indices, ANS Best Estimate Plus Uncertainty International Conference (BEPU 2020), Sicily, Italy, October 2020.

P. Lemaitre, E. Sergienko, A. Arnaud, N. Bousquet, F. Gamboa and B. Iooss, 2015, Density modification based reliability sensitivity analysis, Journal of Statistical Computation and Simulation, 85:1200-1223.

R. Sueur, N. Bousquet, B. Iooss and J. Bect, 2016, Perturbed-Law based sensitivity Indices for sensitivity analysis in structural reliability, Proceedings of the SAMO 2016 Conference, Reunion Island, France, December 2016.

R. Sueur, B. Iooss and T. Delage, 2017, Sensitivity analysis using perturbed-law based indices for quantiles and application to an industrial case, 10th International Conference on Mathematical Methods in Reliability (MMR 2017), Grenoble, France, July 2017.

Examples

Run this code

# NOT RUN {
# Model: 3D function 

  distribution = list()
	for (i in 1:3) distribution[[i]]=list("norm",c(0,1))
  
# Monte Carlo sampling 

  N = 10000
	X = matrix(0,ncol=3,nrow=N)
	for(i in 1:3) X[,i] = rnorm(N,0,1)
     
	Y = 2 * X[,1] + X[,2] + X[,3]/2
	q95 = quantile(Y,0.95)
	
	nboot=200
	
# sensitivity indices with perturbation of the mean 
  
	v_delta = seq(-1,1,1/10) 
	toto = PLIquantile(0.95,X,Y,deltasvector=v_delta,
	  InputDistributions=distribution,type="MOY",samedelta=TRUE,
	  percentage=FALSE,nboot=nboot)

# Plotting the PLI
  par(mar=c(4,5,1,1))
	plot(v_delta,toto$PLI[,2],ylim=c(-1.5,1.5),xlab=expression(delta),
		ylab=expression(hat(PLI[i*delta])),pch=19,cex=1.5)
	points(v_delta,toto$PLI[,1],col="darkgreen",pch=15,cex=1.5)
	points(v_delta,toto$PLI[,3],col="red",pch=17,cex=1.5)
	lines(v_delta,toto$PLICIinf[,2],col="black")
	lines(v_delta,toto$PLICIsup[,2],col="black")
	lines(v_delta,toto$PLICIinf[,1],col="darkgreen")
	lines(v_delta,toto$PLICIsup[,1],col="darkgreen")
	lines(v_delta,toto$PLICIinf[,3],col="red")
	lines(v_delta,toto$PLICIsup[,3],col="red")
	abline(h=0,lty=2)
	legend(0.8,1.5,legend=c("X1","X2","X3"),
		col=c("darkgreen","black","red"),pch=c(15,19,17),cex=1.5)
  
# Plotting the perturbed quantiles
  par(mar=c(4,5,1,1))
	plot(v_delta,toto$quantile[,2],ylim=c(1.5,6.5),xlab=expression(delta),
		ylab=expression(hat(q[i*delta])),pch=19,cex=1.5)
	points(v_delta,toto$quantile[,1],col="darkgreen",pch=15,cex=1.5)
	points(v_delta,toto$quantile[,3],col="red",pch=17,cex=1.5)
	lines(v_delta,toto$quantileCIinf[,2],col="black")
	lines(v_delta,toto$quantileCIsup[,2],col="black")
	lines(v_delta,toto$quantileCIinf[,1],col="darkgreen")
	lines(v_delta,toto$quantileCIsup[,1],col="darkgreen")
	lines(v_delta,toto$quantileCIinf[,3],col="red")
	lines(v_delta,toto$quantileCIsup[,3],col="red")
	abline(h=q95,lty=2)
	legend(0.5,2.4,legend=c("X1","X2","X3"),
		col=c("darkgreen","black","red"),pch=c(15,19,17),cex=1.5)
		
# Plotting the PLI in percentage with refined confidence intervals
	toto = PLIquantile(0.95,X,Y,deltasvector=v_delta,
	  InputDistributions=distribution,type="MOY",samedelta=TRUE,
	  percentage=TRUE,nboot=nboot,bootsample=FALSE)
	  
  par(mar=c(4,5,1,1))
	plot(v_delta,toto$PLI[,2],ylim=c(-0.6,0.6),xlab=expression(delta),
		ylab=expression(hat(PLI[i*delta])),pch=19,cex=1.5)
	points(v_delta,toto$PLI[,1],col="darkgreen",pch=15,cex=1.5)
	points(v_delta,toto$PLI[,3],col="red",pch=17,cex=1.5)
	lines(v_delta,toto$PLICIinf[,2],col="black")
	lines(v_delta,toto$PLICIsup[,2],col="black")
	lines(v_delta,toto$PLICIinf[,1],col="darkgreen")
	lines(v_delta,toto$PLICIsup[,1],col="darkgreen")
	lines(v_delta,toto$PLICIinf[,3],col="red")
	lines(v_delta,toto$PLICIsup[,3],col="red")
	abline(h=0,lty=2)
	legend(0.,1.,legend=c("X1","X2","X3"),
		col=c("darkgreen","black","red"),pch=c(15,19,17),cex=1.5)
  
# }

Run the code above in your browser using DataLab