[R] package "sensitivity" - sobol2002 indices

From: Duarte Viana <viana.sptd_at_gmail.com>
Date: Mon, 18 Apr 2011 18:57:02 +0200


Hello all,

In the following example, one sobol index for one parameter (the first order index for parameter b) is larger than 1. Shouldn´t the indices have a range of [0,1]?

# Code for a decoupled approach example:

X1<-data.frame(a=runif(100,20,30),b=runif(100,0.1,0.5)) X2<-data.frame(a=runif(100,20,30),b=runif(100,0.1,0.5))

library(sensitivity)

x<-sobol2002(model=NULL,X1,X2,nboot=100)

resp<-x$X$a*x$X$b # Example of model output

tell(x,y=resp)

print(x)

# End

Result

Call:
sobol2002(model = NULL, X1 = X1, X2 = X2, nboot = 100)

Model runs: 400

First order indices:

    original bias std. error min. c.i. max. c.i. a 0.08852067 -0.01317529 0.1227359 -0.2126156 0.3132482 b 1.24009414 0.01207408 0.4018650 0.2074935 1.8574080

Total indices:

   original bias std. error min. c.i. max. c.i. a 0.1172252 0.018504397 0.1501931 -0.1456118 0.4896815 b 0.4323300 -0.004814019 0.2692777 -0.1669219 1.0103819

What am I doing wrong? Or the approach has some limitation that I might be disregarding?
Thanks for any clarification on this.

Duarte Viana



R-help_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Received on Mon 18 Apr 2011 - 17:01:15 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Mon 18 Apr 2011 - 17:10:31 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive