Re: random data

From: Mok-Kong Shen (mok-kong.shen_at_t-online.de)
Date: 07/31/03 

Date: Thu, 31 Jul 2003 17:39:02 +0200

Bryan Olson wrote:
>
> Mok-Kong Shen wrote:
>
> > Does one need to apply a K-S test on the p-values or
> > something?
>
> The Kolmogorov-Smirnov test is often a good way to combine the
> results of running the same tests on many independent data-sets.

A presumably dumb question: If I don't err, Diehard
contains a number of different tests, from each of which
p-values are computed. Would it be permissible to lump
all p-values from all kinds of tests in it for doing one
single K-S test (since all p-values are tested against
the uniform distribution)? Thanks.

M. K. Shen

Relevant Pages

  • Re: AES and Diehard
    ... Bryan Olson wrote: ... > Danilo Gligoroski wrote: ... >> begining of the file, all tests in Diehard passed without p values ... With the new version and random keys the p-values ...
    (sci.crypt)
  • Re: AES and Diehard
    ... Bryan Olson wrote: ... > Overlapping sums test. ... How have you got those p-values? ... Cristiano ...
    (sci.crypt)
  • Re: Bootstrapping Hypothesis Test
    ... I really am talking about bootstrapping. ... but not quite equal to the uniform distribution. ... Looking at their p-values also seems odd. ... > - What allows a Monte Carlo estimator of a statistic ...
    (sci.stat.math)
  • Re: Interpreting the test result on a RNG
    ... To check whether the KS p-values are good I ... I expect that real RNGs will tend to fail one particular ... a couple single-test KS scores are quite small. ... smallest few values from a uniform distribution?" ...
    (sci.crypt)
Loading