Hi Rooters, I just finished implementation of a class which uses the Profile Likelihood method to calculate confidence intervals (for a Poisson process with background) of arbitrary confidence level in presense of systematic (and statistical) uncertainties in the signal efficiency and background estimate. You find the class: http://root.cern.ch/root/htmldoc/TRolke.html and it is in the CVS archive since a few days. For the more statistically interested: The method follows the unified scheme (ala Feldman-Cousins) and uses a profile likelihood to treat the nuisance parameters. It has good coverage properties. And it is to the best of my knowledge the only available FULLY frequentist method to treat systematics in confidence interval calculation. A publication is about to be submitted. cheers, Jan Conrad --------------------------------------------------------------------- J. Conrad, AIT-PT,PH-Department, CERN, CH-1211 Geneva 23, Switzerland
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:05 MET