Hi, Claus and others, Here is what I think how we should correct for the missing fraction (~15%) of NSD events in pp. A: Calculate pt spectra from HIJING (NSD) without any trigger conditions at given eta B: Calculate pt spectra from HIJING (NSD) with the INEL trigger condition at given eta M: Measured pt spectra at given eta R: Correction Factor: R(pt) = B(pt)/A(pt) Then, C: Corrected spectra C(pt) = R(pt)*M(pt) We cannot simply multiply 1/0.85 in the spectra, since the bias is nonuniformly distributed in all impact parameter AND in pt. If we want to correct for the missing fraction in NSD in pp, I think what I described above is probably what we have to do. JH > -----Original Message----- > From: brahms-l-bounces@lists.bnl.gov > [mailto:brahms-l-bounces@lists.bnl.gov] On Behalf Of Claus O. > E. Jorgensen > Sent: Wednesday, February 11, 2004 4:30 PM > To: Hironori Ito > Cc: brahms-l > Subject: Re: [Brahms-l] INEL trigger efficiency in pp > > > > Thanks Hiro, > > It's still not completely clear to me what the correct > procedure for this efficiency correction is. > > We measure a number of tracks (Ntr) and a number of events > (Nev) and divide those out to get the yield (Ntr/Nev). We > know that we only pick up a fraction of the total cross > section (72%) so we can divide Nev by 0.72 to get the total > number of inelastic events, but what about the tracks in > these (28%) missing events? How do we evaluate the Ntr in the > missed events. If we just calculate the yields by dividing > Nev by the trigger efficiency (i.e. multiply the spectrum by > 0.72) we somehow assume that the missing events don't have > any tracks. Is that fair? > > I've tried to understand how STAR is doing it. It's not > written explicitly, but this is how I interpret them: They > claim that they measure 85% of NSD events and gives an > uncertainty of 14% for the normalization to their NSD > spectrum - I guess the multiply their spectrum by 0.85 and > the 14% is for the tracks in the missed events (it could in > principle be zero or as abundant as in the measured events). > They also claim that the NSD and total inel yields are almost > identical (only a small difference at low pt) - I'm not sure > I understand that, since I wouldn't expect the single > diffractive (SD) events to contribute with very many tracks > (in their acceptance) but they are still events (and should > go into Nev in the calculated yield). In other words I > wouldn't expect the same yield in SD events as in NSD events. > > I'm a bit confused - who knows about these things and what > should the official BRAHMS procedure be? > > Cheers, > > Claus > > > On Tue, 10 Feb 2004, Hironori Ito wrote: > > > Ok. JH pointed out to me that the option is IHPR2 not > IHNT2. I run > > hijing with non-single diffractive option. Now, it > produces dn/deta > > =2.5 at eta =0. This is right number. Using this HIJING pp in our > > BRAG (GEANT), INEL counter efficiency is now 85%. So, as a > summary, > > this is my result. > > > > INEL efficency > > 72% in total inelastic cross section > > 85% in non-single diffractive cross section > > > > Hiro > > > > Hironori Ito wrote: > > > > > Hello. I guess I made a mistake in efficiency in non-single > > > diffractive events since 10% increase in dn/deta is not > the same as > > > 10% in cross section. As JH suggested, I tried to run > Hijing with > > > non-single diffractive option using IHNT2(13) =3 option. > But, I did > > > not see any change in dn/deta. Therefore, I am not quite > sure how > > > to run Hijing with this option. > > > > > > Hiro > > > > > > Hironori Ito wrote: > > > > > >> Hello. Since so many people asked about INEL efficency, > I just dig > > >> up my files. (I thought these things are checked by someone who > > >> are writing Ph.d thesis. :) ) Here is the conclusion I > made. 1. > > >> From the Hijing 1.383, our INEL counter is about 72% efficient. > > >> (see > > >> > http://www4.rcf.bnl.gov/~hito/run03/hijing_pp_inel_efficiency.gif ) > > >> > > >> 2. Now, looking into what HIJING 1.383 really is for > pp. I looked > > >> at dn/deta from raw hijing output. At, eta=0, it is > 2.2 . From > > >> UA5 ppbar results (see > http://pdg.lbl.gov/2002/contents_plots.html > > >> you can > find a postscrip file there.), it shows 2.2 with the words > > >> saying "The number per pseudorapidity interval is about > 10% higher > > >> if the rate is normalized excluding singly diffractive events > > >> rather than to the total inelastic rate." This tells me > that Hijing pp produces the > > >> total inelastic collisions. (This also means the > following. Since > > >> we don't trigger on single diffractive events, what we > can measure > > >> is 90% efficient at most. Our INEL is 72/90=80 % efficient for > > >> non-single diffractive events.) > > >> > > >> > > >> If people need it, I can also dig my old files for dAu. > > >> > > >> > > >> Hiro > > >> > > >> _______________________________________________ > > >> Brahms-l mailing list > > >> Brahms-l@lists.bnl.gov > > >> http://lists.bnl.gov/mailman/listinfo/brahms-l > > > > > > > > > > > > > > > _______________________________________________ > > > Brahms-l mailing list > > > Brahms-l@lists.bnl.gov > > > http://lists.bnl.gov/mailman/listinfo/brahms-l > > > > > > > > _______________________________________________ > > Brahms-l mailing list > > Brahms-l@lists.bnl.gov > http://lists.bnl.gov/mailman/listinfo/brahms-l > > > > > _______________________________________________ > Brahms-l mailing list > Brahms-l@lists.bnl.gov http://lists.bnl.gov/mailman/listinfo/brahms-l > _______________________________________________ Brahms-l mailing list Brahms-l@lists.bnl.gov http://lists.bnl.gov/mailman/listinfo/brahms-lReceived on Thu Feb 12 10:38:25 2004
This archive was generated by hypermail 2.1.8 : Thu Feb 12 2004 - 10:38:37 EST
