"ZHANG Yan" <[EMAIL PROTECTED]> wrote:

> if the distributions of n indepednet random variables Y1,Y2,...Yn 
> are given, ie. P(Yi=yi) is known for i=1..n;
> 
> I need to find the following probability:
> P(Y1+Y2+...+Yn = M)=?

The density for the sum of independent variables is the convolution 
of the density of each variable. This applies to both continuous and 
discrete variables.

That is, if you have density functions p1, p2, ..., pn, then the
density of the sum is p1 * p2 * ... * pn, denoting the convolution
with the asterisk.

The convolution is more quickly calculated by appealing to the
relation FT(f*g) = FT(f) FT(g), that is, the Fourier transform of
the convolution is the product of the Fourier transforms. This is
generally much, much faster than computing the convolution via
summation or integration.

I use Octave (http://www.octave.org) for calculations of this kind.

For what it's worth,
Robert Dodier
--
``If you could count for a year, would you get to infinity,
Or somewhere in that vicinity?'' -- Tom Lehrer
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to