[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Normally distributed vectors

I sent by mistake my reply just to Ewan Macpherson, instead of the whole
newsgroup. However, since it might be of interest to other people, I post it
here together with his reply.

> > > 
> > > All you need to do is generate each component from
> > > independent normal distributions.
> > > 
> > 
> > Hmmm... I might be wrong, but I think that this is true only if the
> > covariance matrix (the 3X3 matrix made up of variances and
> > covariances) is diagonal. In the other cases, I should find the
> > basis where it is diagonal, sample from independent normal
> > distributions, and then go back to the original basis. Just some
> > linear algebra.
> Yes indeed.  I misread the part of your message where you 
> specified NON-zero covariances.  Oops!
> cheers,
> --
> Ewan Macpherson   <emacpher@umich.edu>
> Central Systems Laboratory
> Kresge Hearing Research Institute
> http://www-personal.umich.edu/~emacpher

So, thanks Ewan for your attempt.  Of course, my original request is still
valid. Greetings to everyone.