Re: Lognormal distribution

[Jean Bréfort <jean brefort normalesup org>]

Le vendredi 22 août 2008 à 07:07 -0600, Andreas J. Guelzow a écrit :
> On Fri, 2008-08-22 at 12:07 +0200, Jean Bréfort wrote:
> > Hi,
> >
> > Our Lognormal distribution support does not have a shape parameter as
> > described at
> > http://www.itl.nist.gov/div898/handbook/eda/section3/eda3669.htm
> >
> > Is that correct or a gnumeric bug (or a R-project bug)?
>
> Note from the same page:
> Since the general form of probability functions can be expressed in
> terms of the standard distribution, all subsequent formulas in this
> section are given for the standard form of the function.
>
> So both the m (shape parameter) and theta are unnecessary. We apparently
> chose to include theta but not m.
>
> SO I don't see a bug anywhere.

Hmm, theta might be <0 on the page, but not our parameters, so the
various parameters have apparently not the same signification. I must
think a bit more about that.

Jean