NORTARA (version 1.0.0)

nortaRA-package: Simultaneous generation of multivariate data with arbitrary marginals

Description

This package implements a specific method for generating n-dimensional random vectors with given marginal distributions and correlation matrix. The method uses the NORTA(NORmal To Anything) approach which generates a standard normal random vector and then transforms it into a random vector with specified marginal distributions and the RA(Retrospective Approximation) algorithm which is a generic stochastic root-finding algorithm. Data generation is accomplished by first using BoundingRA to calculate an intermediate multivariate normal correlation matrix, then the matrix is used to generate samples from multivariate normal distribution. The engine function genNORTARA will transforms the normal samples to the wanted data set with specified inputmarginals from users. The function valid_input_cormat returns the lower and upper bounds of the mixture pre-specified marginal distributions. The function check_input_cormat checks the input target correlation matrix whether it is in the lower and upper bounds, if in the bounds, then the function will return TRUE it means the input target correaltion matrix is feasible, otherwise, it will print the elements' positions which are out of bounds and give an error message.

Details

Package:

nortaRA

Type:

Package

Version:

1.0.0

Date:

2014-12-06