Summary of STAR HBT PWG phone meeting 2 Sept 99 Attending from BNL, OSU, UW, UT, LBL: Selemon Bekele, John Cramer, Fabrice Fretiere, Tom Humanic, Frank Laue, Mike Lisa, Sergei Panitkin, Lanny Ray, Dennis Reichhold, Fuqiang Wang, Randy Wells, Robert Willson, Jingbo Zhang -------------------------------------------------------- Randy Wells has implemented a StHbtCoulomb class in the StHbtMaker framework, and he gave an update. It has several methods for returning the Coulomb correction for a pair of charged particles. The correction is based on a detailed numerical integration of Coulomb wavefunctions that John Cramer has done for like-sign and opposite-sign particles. John has put his calculations for spherical sources of various radii (R=0..20 fm, in 2 fm steps) in the STAR HBT afs area. You specify the radius you want (it need not be one of the ones that John calculated for), and a StHbtCoulomb object reads John's calculations file (or another file of calculations, if you use the SetFile() method of the object), and creates a lookup table of corrections in eta. This lookup table is built by a linear interpolation in radius. Then, when one requests a Coulomb correction, another linear interpolation in eta is performed from the lookup table. If you specify the radius in a call, the object will check to see if it needs to update its eta lookup table, and will do the update if neccessary. The class is intended to be as "foolproof" as possible. Randy sees this class as essentially a finished product (although Fuqiang Wang has pointed out a likely "factor of 2" problem in the eta calculation from Qinv). For more details, Randy has posted web pages (hung from the "Projects" section of the STAR HBT web page). > John Cramer pointed out that the R=0 fm Coulomb wave integration, which should be identical to a Gamow function, *is* essentially a Gamow, but has numerical noise. He suggests using Gamow instead. > Randy says he will use the Gamow for R=0, and wonders if the R=2 fm calculation could also be redone to eliminate noise. -------------------------------------------------------- Helen Caines and Tom Humanic are working to incorporate V0's into the StHbt framework. At our Brookhaven meeting, this was identified as a priority. The idea is to build on the work of the strangeness group. They have developed a class for their microDST, called StV0MiniDst. They fill a collection with objects of this class. The HBT plan is to run our maker *after* theirs in the chain, and simply copy into the StHbtEvent the information in their object. One complication is that the StV0MiniDst inherits from TObject. Since we would like to be "T free", we will copy the information into a StHbtV0 class. The StHbtEvent will have a collection of these objects (StHbtV0Collection) just like it now has a collection of StHbtTracks for charged particles. For doing V0-V0 correlations, it is forseen to write a StHbtEventReader class that will read directly the strangeness group's microDSTs. This should make V0-V0 HBT extremely efficient. However, *all* StHbtEventReaders (currently, there is only the StStandardHbtEventReader which reads from StEvent, and the StHbtMcEventReader which reads from StMcEvent) will be updated to fill the StHbtV0 structures. Helen reported that the status of this project is that she is able to get a pointer to the collection created by the strangeness group's maker. Next, she should be able to get a pointer to the StV0MiniDst objects themselves, and then copy the information into the StHbtV0 objects. She and Tom expect a useful first release in 2-3 weeks. > Fuqiang pointed out that it would be very useful to have an estimate of the expected background (from "false" V0's) experimentally. Therefore, we should not make too tight a cut on the invariant mass of the V0. > Helen replied that the microDST does not have a very strict cut on the V0 mass. The strict cut will be done in the StHbtAnalysis by a StHbtV0Cut. The spectrum for particles that pass the cut, as well as those which fail the cut, will be kept. From this, an estimate of the background can be extracted. > Dave Hardtke asked whether this replaced his StHbtV0Track that he developed as a placeholder, based on the HBT PWG meeting at BNL in July. > Yes. > Dave pointed out that his StHbtV0Track class had pointers to the daughter tracks. Is something like that available in the StV0MiniDst? > No. > Dave and Fuqiang mentioned that we should be careful to keep enough information about the daughter tracks to be able to do two-track cuts, and to extract information about the decay angle. > Helen pointed out that two-track seperation is not a big issue in V0 interferometry, and that much information about the daughters is kept in the StV0MiniDst object itself. The decay angle, for example, can be extracted >>>>> REQUEST TO EVERYONE: Please look at the information available in the proposed StHbtV0 class, and make sure it is sufficient. A web page with the proposal will be linked to the "Projects" area of the STAR HBT page. > Dave pointed out that Lambda-Lambda HBT has been done at LEP. We should see what information they found neccessary to keep. He will send the paper to the hbt email list. -------------------------------------------------------- Robert Willson reported that he has a gotten a fitting class from Dave Hardtke. He plans to work on integrating it into the StHbtMaker framework, and figure out how to do log-likelihood as well as chi-squared based fits. It uses minuit called from root. -------------------------------------------------------- Fabrice Fretiere has been quite busy with TPC calibrations recently, but will work on incorporating "hidden" information (like a particle's true momentum and species, if it is Monte Carlo information) into the StHbtEvent objects. He had originally envisioned 2 LorentzVectors, but now imagines that it will be a lot more. Is this a problem in terms of space? > Mike: no, it is not a problem. The hidden information object will be incorporated into the StHbtTrack class (and StHbtV0 class, when it exists) only as a member pointer. So, if one chooses not to have the "hidden" information, then only the space of a single pointer is wasted. -------------------------------------------------------- Randy has found that the MDC2-era files are becoming impossible to read with dev software. This includes the DST files for Lanny Ray's correlated pion events. Mike and Randy will choose 2-3 useful files, and ask Lidia to re-process them though the BFC, making new DST files. > Lanny Ray had to leave the meeting before we discussed this, but later said that the new DST tables and StEvent would take some time to settle down. He suggests proceeding as best we can, and not wait for the new DST/StEvent. -------------------------------------------------------- Mike Lisa plans to implement a persistent HBT microDST. It will be ASCII-based, and should be readable on all computer platforms. The I/O of these files will NOT depend on root, so a non-root-based version of the StHbt framework will be possible. He expects to have a working prototype in about a week. -------------------------------------------------------- The relative brevity of the meeting (40 min) was appreciated by all, and we will try hard to keep it that way. 60 minutes will remain the strict maximum. For now, we expect to continue the phone meetings at the same time: every other Thursday at noon Eastern time. - summary by M. Lisa