Command:

setCorrelationStein (  
int numFactors
int numSamples
double** corrSample
)

What:

Sets the correlation method between factors to Stein method. This method is useful when the user is given an empirical correlated sample and s/he wants to generate a Latin Hypercube Sample with the same correlation structure of the given sample.

Parameters:

numFactors
[integer]
the number of factors (e.g. columns in the sample matrix)
numSamples
[integer]
the number of samples (e.g. rows in the sample matrix)
corrSample
[matrix of double]
the correlated sample to start from. It must be a matrix NxM with N the number of samples and M the number of factors

Return:

error_code [int]: 1 if the operation has occurred without errors, 0 if an error has occurred

Examples:
% sampleIman is a given empirical NxM sample matrix
setCorrelationStein(3,10,sampleIman)
sampleStein=createSample

this command in Matlab returns in the variable sample an NxM matrix containing the generated sample, where N is the number of runs and M the number o input variables.


About the correlation method:

In order to use the Stein (see [23]) methods the user provides a sample matrix that contains a correlated sample (in the form of, e.g, a random sample, or an empirical sample obtained by an experiment). The method generates an LHS (or a replicated-LHS) sample that has the same correlation of the sample provided by the user. 

All Correlation Commands: