Hi Eiko, can't you make a profile histogram of d0 in terms of phi0 bins and then fit d0(phi0) data points with Acos + Bsin? Or you really want an unbinned likelihood fit of data points? In that case, if the errors are gaussian and uncorrelated, it boils down to the profile histogram, and if they are not, it's going to be very hard to write a probability function? Cheers, Petar
This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:38 MET