PCESI: Compute Sensitivity Indexes from a PCE Design

In polychaosbasics: Sensitivity Indexes Calculated from Polynomial Chaos Expansions

Description

Calculation of PCE sensitivity indexes and related results.

Usage

PCESI(poly)

Arguments

poly	an object of class PCEpoly. The design to analyze (a Legendre polynomial).

Value

An object of class PCEfit.

Note

• By default, only a part of the returned object is displayed by the functions print() and show(). To see the hidden components, use the function print() with the option all=TRUE or the function getNames (see methods of class PCEfit).

• It is adviced to increase gradually the polynomial degree, up to the returned object contains a R2 value near from 1 and a low RMSEP value.

See Also

• Functions polyLeg and analyticsPolyLeg, creators of objects from class PCEpoly.

• Class PCEfit for description of the returned structure.

Examples

# Dataset simulated by  using the Ishigami function 
nlhs <- 200 # number of rows 
degree <- 6 # polynomial degree
set.seed(42)# fix the seed for reproductible results
pce <- analyticsPolyLeg(nlhs, degree, 'ishigami')# build the PCE design
ret <- PCESI(pce) # compute the PCE sensivity indexes
print(ret, all=TRUE)

polychaosbasics documentation built on May 29, 2017, 12:58 p.m.