help ajv                                                 Version 1.1 2007-02-26
-------------------------------------------------------------------------------

Title

ajv -- Generate a random variate following a Johnson distribution

Syntax

ajv [varname] [if] [in] , distribution(SN|SL|SU|SB) generate(newvar) gamma(#) delta(#) [options]

options Description ------------------------------------------------------------------------- Main lambda(#) value for lambda; defaults to one xi(#) value for xi; defaults to zero seed(#) seed for random number generator n(#) set number of observations -------------------------------------------------------------------------

Description

ajv generates a random variable newvar that follows one of the Johnson distributions, SN, SL, SU or SB. Parameters gamma, delta, lambda and xi are such as those obtained from jnsn. If varname is specified, then it will be used as a standard normal deviate and newvar will be generated from it and the Johnson distribution parameters. If no varname is specified, then ajv will internally generate a standard normal variate using the seed() if one is specified.

Options

+------+ ----+ Main +-------------------------------------------------------------

distribution specifies the Johnson distribution; it may be in upper, lower or mixed case.

generate names the new variable.

gamma delta lambda xi specify parameters of the Johnson distribution.

seed sets the seed for random number generation, e.g., when varname is not specified.

n specifies the number of observations, e.g., starting from an empty dataset.

Remarks

ajv is an implementation of Algorithm AS 100.1 (Hill, 1976).

References

I. D. Hill, Normal-Johnson and Johnson-Normal transformations. Applied Statistics 25:190–92, 1976.

Examples

. ajv snv, distribution(sl) generate(lnv) gamma(1) delta(1)

. ajv , distribution(su) generate(suv) gamma(1) delta(1) seed(12345) n(100)

Author

Joseph Coveney jcoveney@bigplanet.com

Also see