Re: [ROOT] Low Statistics Feldman Cousins

From: Rene Brun (Rene.Brun@cern.ch)
Date: Mon Feb 18 2002 - 09:14:17 MET


Hi Kevin,

If you search with Google you will find several references on this subject:
Search for "Low Statistics Feldman Cousins"

If somebody is willing to provide a working function, I will add it to TMath
or equivalent.

Rene Brun

Kevin Reil wrote:
> 
> Hi,
> 
> hi=FeldCous->Upper(68.27,2)             (hi=4.25)
> lo=FeldCous->Lower(68.27,2)             (lo=0.74)
> 
> Does anyone have a function which can return confidence intervals
> IE replicate at 68.27% confidence
> 
> 0 (0.00-1.29) instead of (0.00-0.00) from sqrt(n)
> 1 (0.37-2.75)            (0.00-2.00)
> 2 (0.74-4.25)            (0.59-3.41)
> 4 (2.34-6.78)            (2.00-6.00)
> 9 (6.33-12.79)           (6.0-12.00)
> etc.
> 
> Thanks,
> Kevin
> 
> PS an electronically accessible reference would also help. I have these
> numbers from a document and cannot get to reference.
> 
> |---------------------------------|------------------------------------|
> | Campus Address                  | Office Location                    |
> | 115 S, 1400 E, Room 201         | 262 INSCC                          |
> | Salt Lake City, UT 84112-0830   | Just North of Park Building        |
> |---------------------------------|------------------------------------|
> | www.physics.utah.edu/~reil      | Office (801) 581-4306              |
> | reil@physics.utah.edu           | Home   (801) 969-6399              |
> | kevin@reil.net                  | Fax    (801) 581-6256 (via Frank)  |
> |----------------------------------------------------------------------|
> |                    And my father dwelt in a tent.                    |
> |----------------------------------------------------------------------|



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:42 MET