1 Hough-4-14

2 Outline Solve memory leak problem The D->Kpipi from last weekTracking efficiency with combined tracking without inner chamber First try the easy method on stereo wire

3 Memory leak problem in dataValgrind to check Track pointer not delete in root fitting A whole raw file reconstructing won’t fail Submit one week data to reconstruction(not finish) One day’s data has not enough low pt events

4 D->kπ π Last week work by checkSomething error in analysis code ->make 100% efficiency of D one pattsf.txt job is failed in background , make it 5% signal lose After fix these problem , nearly signals by PATTSF & PATTSF+HOUGH(~50% efficiency ,coincide with physics report) Check track momentum , not much low pt track in D->k π π A track pt distribution from D->kpipi GeV GeV

5 Check with no inner chamberData sample: MC single muon fixpt 10,000 events Non uniformed magnetic Mix background in reconstruction Reconstruction method: PATTSF+HOUGH PATTSF

6 Check with no inner chamberEfficiency definition: events at least find 1 track/all events (above 250MeV) , open to reconstruct higher pt for Hough) 50MeV 75MeV 100MeV 150MeV 250MeV 500MeV 1GeV PATTSF ~0 ~1% 41% 93.7% ~95% ~96% PATTSF+HOUGH ~2% 52% 86% 95% Hough now hardly take back higher pt without inner chamber

7 Method to deal with stereo hitsFrom big angle track (|cos|>0.8) , 2d fitting performs good , most lost track are due to stereo hits loss We can use 2D information to collect stereo hits that near to the circle track some cut condition is under studying