Hi Eugene,
I can see, that track fitter rejects a lot of tracks in your case. You can try to turn
off fitting - it defenitely increases the number of found tracks to the level of more
than 30 tracks/event in current version of track finder (Mike Ronan has improved
version, which should do even better, but this is not in recent release, and it will work
much slower, in fact). How to turn off fitting is described in Mike Ronan letter to
David Wagner from April 21, and it was send to everybody in lcd-sim list,
so you probably have it. If you don't, I can forward it to you.
I don't know, if it will help with fraction of the tracks, matching MC generated ones.
Without fitting it may be even worse, but if fitter mulfunctionning - who knows?
Nick Sinev
Eugene Guillian wrote:
> Hi,
>
> I've been running hep.lcd.recon.tracking.TrackReco on ttbar events
> simulated for the large detector, and I feel uncomfortable about the
> results I've been getting. I wonder if somebody could shed some light
> as to what I'm doing wrong, or whether what I'm seeing isn't unexpected.
>
> I'm focusing my study on ttbar events in the "lepton + jets" decay
> channel -- i.e. one of the top quarks decays to ele or muo + neutrino +
> b-quark, the other to three quarks. When I run TrackReco, I am able to
> match the primary muon track to the monte carlo track in only 58% of the
> events; with electrons, this figure goes down to 31% (check out plots in
> umaxp1.physics.lsa.umich.edu/~guillian/nlc/23-apr-1999/summary.html).
> What's more, the average number of charged monte carlo tracks is 45.0,
> while the number of found tracks is 11.6. These figures seem rather
> dismal. Is this expected? Or do the default track-finding paramters
> need to be tuned for ttbar events?
>
> Gene
|