[ROOT] KS - Kolmogorov-Smirnov

From: Stanley Forrester (sforrest@lifshitz.ucdavis.edu)
Date: Tue May 21 2002 - 17:52:39 MEST


I would like to compair two histograms with the Kolmogorov-Smirnov test or 
some similar goodness of fit test.  I know that chi squared and RMS are 
both avaliable as standard tests.  Are either the K-S or Anderson-Darling 
avaliable in root.  If so how do I access them?

-- 
Stanley S. Forrester
sforrest@lifshitz.ucdavis.edu


---------- Forwarded message ----------
Date: Mon, 06 May 2002 13:02:30 -0500
From: Alexei N Safonov <safonov@fnal.gov>
To: "Sforrest@Fnal. Gov" <sforrest@fnal.gov>
Cc: sforrest@fnal.gov, jrsmith@fnal.gov, Maxwell Chertok <chertok@fnal.gov>
Subject: RE: Your mail and Progress

KS - Kolmogorov-Smirnov test. It is standard in root.
I presume you will have to look around a bit, but it
should be not too difficult to find. If needed, we can
look later together. For now, I suggest we produce the
two files and then discuss how to compare them. I will
ask Rob to send us his script and we will take a close
look.

Cheers,
Alexei

-----Original Message-----
From: Stanley Forrester [mailto:sforrest@lifshitz.ucdavis.edu]
Sent: Monday, May 06, 2002 12:50 PM
To: Alexei N Safonov
Cc: sforrest@fnal.gov; jrsmith@fnal.gov; Maxwell Chertok
Subject: RE: Your mail and Progress


> I suggest you take a few another runs and produce 
> the root files (take runs from a different time range). 
> We will later need to work out the procedure for comparing
> them to the sample one (I suggest to use KS test).

I'll continue using jbel01.  I'll have to look up how to tell different 
time ranges and could you explain what a KS test is?



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:53 MET