[SciPy-User] log normal distribution random number array generation

[Robert Kern]

On Tue, Dec 3, 2013 at 11:08 PM, <josef.pktd at gmail.com> wrote:
>
> On Tue, Dec 3, 2013 at 5:46 PM, Paul Hobson <pmhobson at gmail.com> wrote:
> > Jose,
> >
> > For lognorm.rvs, mu and sigma translate to loc and scale, respectively.
The
> > same is true for norm.rvs
>
> For the lognorm, mu and sigma are often used as parameters of the
> underlying normal distribution, not directly of the lognormal mean and
> scale
>
> "If log(x) is normally distributed with mean mu and variance sigma**2,
> then x is log-normally distributed with shape parameter sigma and
> scale parameter exp(mu)."
>
http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.lognorm.html

Specifically, in order to translate any standard convention to lognorm, you
must keep the default loc=0. Most standard conventions for the log-normal
distribution do not shift the location at all, just the scale and a shape
as explained above.

--
Robert Kern
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20131204/0543536e/attachment.html>