Y = (X - xi) / lambda
that a unit normal Z can be represented as
gamma + delta * logit Y
So, mixing in Stata functions to the algebra,
Y = invlogit((Z - gamma)/delta)
X = xi + lambda * invlogit((Z - gamma)/delta)
So it looks like
. gen XS_B = xi + lambda * invlogit((invnorm(uniform()) - gamma)/delta)
so long as you plug in numbers for the parameters
xi, lambda, gamma, delta. Here xi is the minimum and lambda is
the range.
Check my algebra!
Nick
[email protected]
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Sascha O.
> Becker
> Sent: 20 July 2004 10:08
> To: [email protected]
> Subject: st: Johnson's SB distribution
>
>
> Dear all,
>
> some months ago there was a mail exchange on statalist
> mentioning Johnson's
> distributions. I could not infer from this discussion if
> there is any Stata
> module around that allows to draw random numbers from Johnson's SB
> distribution? A -net search- and googling was not successful in this
> respect.
> Any ideas on how to do this? I wanted to replicate a
> simulation study using
> Johnson's distribution (this study was done using some other
> package, i.e.
> not Stata)
>
> Best regards, Sascha
>
> Sascha Becker
> University of Munich, Germany
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
