[Biopython] Statistical similarity in microarray data

[Istvan Albert]

On Tue, Feb 16, 2010 at 8:37 AM, Peter Saffrey <pzs at dcs.gla.ac.uk> wrote:

> that are different samples. It also seems to lack precision. I get p-values
> of 0 quite a lot; even 1e-80 would reassure me that something is really
> happening underneath.

Hello,

Getting 0 for p value does not mean it lacks precision, only that the
value is too small to be computed precisely, hence for all practical
purposes the chance of the null hypothesis being true is zero. Think
of it as a very tiny p-value, one that is so small that it cannot even
be distinguished from zero.

Once numbers are very small the internal representation errors for the
floats is likely to be larger than the claimed p-values.

best,

Istvan
-- 
Istvan Albert
http://www.personal.psu.edu/iua1