Date: Dec 29, 2012 3:42 PM
Author: Luis A. Afonso
Subject: Re: log normal?
Since 1958 (Box-Muller), 50 years ago, people gave up to simulate normal deviations by adding uniform r.v.. This is how much Richard Ulrich is out-of-date. Meanwhile he persists to misdirect people in Statistics regardless our corrective /advisory remarks.
en.wikipedia.org/wiki/Box?Muller_transform - 41k
Luis A. Afonso
Re: log normal? Posted: Dec 27, 2012 2:11 PM
In detail with the algebra. Normal N comes from U1 + U2 + ... + Uk . Log normal -log(N) comes from log(U1) + log(U2) + ... + Uk whichh implies N comes from U1 * U2 * ... *Uk. There may be differing weights for the different Uniform measures being added, but none of them overly influence it. -- Rich Ulrich