ATmet (version 1.2.1)

ATmet-package: Advanced Tools for Metrology

Description

Several functions for smart sampling and sensitivity analysis for metrology applications, including computationally expensive problems.

Details

The ATmet package implements sensitivity analysis functions for metrology applications

The function for smart sampling implements the Latin Hypercube Sampling (LHS) method using the lhs package. The functions for sensitivity analysis implement the Standardized Rank Regression Coefficient (SRRC) and the Sobol' sensitivity indices using the sensitivity package. These methods can be used for computationally expensive problems.

References

I.M. Sobol, S. Tarantola, D. Gatelli, S.S. Kucherenko and W. Mauntz, 2007, Estimating the approximation errors when fixing unessential factors in global sensitivity analysis, Reliability Engineering and System Safety, 92, 957-960.

A. Saltelli, P. Annoni, I. Azzini, F. Campolongo, M. Ratto and S. Tarantola, 2010, Variance based sensitivity analysis of model output. Design and estimator for the total sensitivity index, Computer Physics Communications 181, 259-270.

R. Stocki, 2005, A method to improve design reliability using optimal Latin hypercube sampling, Computer Assisted Mechanics and Engineering Sciences 12, 87-105.

See Also

lhs

sensitivity

Examples

# NOT RUN {
# **********************
# Smart sampling method
# **********************

# }
# NOT RUN {
N<- 100
  k<- 4
  x<- list("X1","X2","X3","X4")
  distrib<- list("norm","norm","unif","t.scaled")
  distrib.pars<- list(list(0,2),list(0,1),list(20,150),list(2,0,1))
  LHSdesign(N,k,distrib,distrib.pars,x)
# }
# NOT RUN {
# **********************
# Sensitivity analysis
# **********************
  
##Simulate the input sample
M=10000
Xmass <- data.frame(X1 = rnorm(M, 100, 5e-5),
                    X2 = rnorm(M, 0.001234, 2e-5),
                    X3 = runif(M, 1100, 1300),
                    X4 = runif(M,7000000,9000000),
                    X5 = runif(M,7950000,8050000))#Data-frame

#Define the measurement model (GUM-S1, 9.3)
calibMass <-function(x){
  return(((x[,1]+x[,2])*(1+(x[,3]-1200)*(1/x[,4]-1/x[,5]))-100)*1e3)
}

#####   Use SRRC with a model function    #####
#Apply sensitivityMet function to evaluate the associated SRRC indices
S_SRRC=sensitivityMet(model=calibMass,x=Xmass, nboot=100, method="SRRC", conf=0.95)
##Print the results
#First order indices
S_SRRC$S1

#####   Use Sobol with a computational code   #####
#Creation of the design for the computation of Sobol sensitivity indices
S_Sobol=sensitivityMet(model=NULL,x=Xmass,y=NULL, nboot=100, method="Sobol", conf=0.95)

#Obtain the design of experiment to submit to the code
XDesign=S_Sobol$SI$X

#Run the computational code with XDesign as a sample of the input quantities
#We use calibMass function (see GUM-S1) as an example
YDesign=calibMass(XDesign)

#Run the Sobol indices calculations with the outputs of the code
S_Sobol$SI=tell(x=S_Sobol$SI,y=YDesign)

##Print the results
#First order indices
S_Sobol$SI$S
#Total order indices
S_Sobol$SI$T 
# }