TRF++ results 17mar98

24mar98 1200

D0 set a a February 1998 milestone for its run II software. This included generating and propagating Monte Carlo tracks, using the tracks to create scifi signals, clustering these signals and finding tracks in the scifi detector. All of this was done in the D0 run II software environment (SRT, edm, d0om, etc.). Here we report the track finding results from that study.

Results here may be compared with the January 98 study.

The results described here were generated with code very similar to that released in TRF++ version 0.40 using the executable findcyl.

Geometry

The detector model was similar to that of January 1998 except the stereo angle was fixed at 2 degrees instead of the fixed pitch 0.001 rad/cm in the latter.

Track generation

Tracks were generated for three types of events:

1000 single muons with 1.0 < pT < 10 GeV/c,
10 t-tbar and
1000 psi->mu mu events with pT of the psi=3GeV.

Cluster/hit simulation

The tracks are propagated through the detector using GEANT assuming a uniform field of 2.0 Tesla and including interactions such as multiple scattering and dE/dx losses.

The detector response and clustering are done using the D0 CFT packages. There is no detector ineffficiency or noise.

Track finding

Track finding was done with TRF++ 0.41. Track finding starts in the outermost phi-layer moving inward through the first four phi-measurements and then proceeds from the outermost stereo layer inward. Tracks are required to cross all 16 surfaces. The algorithms and their parameters may be found in trftest/SFPathBuilder.

Analysis

The track matching and standr ntuple are as described for January 1998.

Results

Results are contained in the standard ntuple files:

/projects/726/mc_1/cft/1000singlemuons_trf.hbk
/projects/726/mc_1/cft/10ttbar_trf.hbk
/projects/726/mc_1/cft/cc_3gev_et_psimumu_000_1k_trf.hbk

The are analyzed with macros in package trftest (soon to be moved to trfcyl). Here are the plots:

singe muon: efficiency , chi-squares , resolutions , significances .
t-tbar: efficiency , chi-squares , resolutions , significances .
psi->mu mu: efficiency , chi-squares , resolutions , significances .

Conclusions

We are able to generate and reconstruct the scifi part of the D0 detector in the D0 software environment. Much work remains.

Future plans

Extend to Si, muon, etc.
Optimize, optimize, optimize.
Account for multiple scattering and dE/dx losses.
And much more.

March 24, 1998 update

The above results did not account for the charge sign of the Monte Carlo tracks. This does not affect the single muons which were all generated with positive charge but does affect the other two samples. Here are corrected reults for the t-tbar sample:

t-tbar with correct MC charge: efficiency , chi-squares , resolutions , significances .

Please send questions or comments to adams@physics.rice.edu or ngraf@fnal.gov .