Container of the results of PCE sensitivity indexes computation.
Objects from the Class
Objects from this class are created by calls to function
matrix with as many rows as inputs and 3 columns. Values of the PCE sensitivity indexes. The row labels are the inputs numbers. The column labels are LE, PE, TPE.
indexes[i, "LE"]is the Linear Effect of the input
indexes[i, "PE"]is the Polynomial Effect (called "SU" in Sudret, 2008). It is the effect of the monomials in which only the input
indexes[i, "TPE"]is the Total Polynomial Effect (often called "SUT"). It is the effect of all the monomials in which the input
matrix. Percentages of the PCE sensitivity indexes, i.e values of
indexesexpressed as percentages of the sums of their columns.
vector of length 2. The values of R2 and RMSEP (RMSEP: Root Mean Square Error Prediction).
vector of length equal to the number of monomials. Individual monomial sensitivity indexes.
vector of length equal to the number of monomials plus one. Regression coefficients. The first one is the constant term.
vector of length equal to the number of rows of the dataset. Metamodel outputs.
object of class
PCEdesign. Matrix coding the polynomial structure.
expression of class ‘call’. The command which creates the
PCEpolyobject used as input in the creator command.
signature(object = "PCEfit"): display the names, class and length of all the components. See the description of the generic function
signature(object = "PCEfit", all=FALSE, ...): method of function
allis set to FALSE (the default), only the components
fitare printed. The additional arguments are passed to the
signature(object = "PCEfit"): same as function
Global sensitivity analysis using polynomial chaos expansions. Bruno Sudret. In Reliability Engineering and System Safety, Vol. 93, Issue 7, July 2008, pages 964-979.
PCESI, creator of objects from this class.
Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.