Re: Chi2Test

> Could you comment on the relative merits/drawbacks of the two approaches.
> I will add your comment in the KolmogorovTest functions.

Kolmogorov approach is based on theory of comparison of probability distribution.
Comparison of histograms is a particular case. Kolmogorov test is independed of form
of distribution and statistics. Of course, bigger statistics as always is better.

Chi2 test is good assimptotically, it is based on errors in bins. If in each bin we have many entries
So error is well defined, both methods give aproximately the same result. But let define number of bins is much bigger than entries. So in each bin we have 0 or 1.
For Kolmogorov test it is irrelevant, but Chi2 test will not work at all. So for not too big statistics Kolmogorov definetely better.

Victor

Victor M. Perevoztchikov perev_at_bnl.gov Brookhaven National Laboratory MS 510A PO Box 5000 Upton NY 11973-5000 tel office : 631-344-7894; fax 631-344-4206;

• Original Message ----- From: "Rene Brun" <brun_at_pcroot.cern.ch> To: "Jan Conrad" <Jan.Conrad_at_cern.ch> Cc: <roottalk_at_pcroot.cern.ch> Sent: Thursday, August 11, 2005 2:46 AM Subject: RE: [ROOT] Chi2Test

> Hi Jan,
>
> On Wed, 3 Aug 2005, Jan Conrad wrote:
>
> > Dear Root-developers,
> >
> > two suggestions:
> >
> > 1) it might be nice to have Chi2Test also return the Chi2/ndf (many
> > people judge comparability
> > directly on that figure). (Not only by the "P" option).
>
> I have implemented your suggestion in TH1::Chi2Test and TH2::chi2Test.
> I have added two new options:
> // "Chi2" : the function returns the Chisquare instead of the
> probability
> // "Chi2/ndf" : the function returns the Chi2/ndf
> >
> > 2) having Chi2Test in place, maybe it is time to get rid of the
> > Kolmogorov-Smirnov Test for
> > histograms ?
>
> I dont't think that users will appreciate this proposal.
> Could you comment on the relative merits/drawbacks of the two approaches.
> I will add your comment in the KolmogorovTest functions.
>
> Rene Brun
>
> >
> > Best,
> > Jan
> >
> >
> >
> >
>
Received on Thu Aug 11 2005 - 18:34:02 MEST