Presentation is loading. Please wait.

Presentation is loading. Please wait.

Update on the new reconstruction software F. Noferini E. Fermi center INFN CNAF EEE Analysis Meeting - 15/06/20151.

Similar presentations


Presentation on theme: "Update on the new reconstruction software F. Noferini E. Fermi center INFN CNAF EEE Analysis Meeting - 15/06/20151."— Presentation transcript:

1 Update on the new reconstruction software F. Noferini E. Fermi center INFN CNAF EEE Analysis Meeting - 15/06/20151

2 outline Software repository, requirements and platform supported Documentation  will be provided soon The c++ structure of the code  classes Main changed with respect to the previous version of the code Discussion: how to proceed EEE Analysis Meeting - 15/06/20152

3 Requirements Linux machine ROOT should be installed (in principle any version is fine) Windows ROOT should be installed (the same version used by us to compile the code because we will provide directly the executables obtained with Visual Studio 10) EEE Analysis Meeting - 15/06/20153

4 Software repository The new code will be available under github in the same repository of the current analyzer. https://github.com/centrofermi/EEE_Analyzer Currently the version under test is in the «dev» branch: Windows executable (under bin dir) EEE Analysis Meeting - 15/06/20154

5 Running under windows 8 The code developed in a linux machine is able to run also on windows (tested with root v5.34.24). The executable will be available in the github repository (compiled with Visual Studio). The only requirement is the root installation (the installer for the proper root version will be provided). Reconstruction can be trigger in windows also as a double click in the.bin file. EEE Analysis Meeting - 15/06/20155

6 The structure Basically the new software is organized in c++ classes to have a more modular structure and to benefit of ROOT functionalities. However the output is written via simple ROOT trees (as in the current version of the Analyzer). The main differences in the output is the fact that the number of tracks per event is not limited. EEE Analysis Meeting - 15/06/20156

7 Event time The event time will be computed in the same way as before (reported in the documentation). The only difference is that now the total TDC counts are computed in the current second and not in the one before. Nanoseconds are extracted from here, f x 10 9, and then corrected for other effects. TDC counts in the two TDCs for the current event TDC counts in the two TDCs in between two GPS events EEE Analysis Meeting - 15/06/20157

8 The Tracker 1.It selects hits in both sides of each chamber and it fills arrays with hit candidates for tracks: strip#[nHit#], deltat#[nHit#], timeAv#[nHit#] where #=,T, M, B (Top, Middle, Bottom chambers). In case two hits are close in space they are clusterized. 2.A pre-selection of candidate tracks is performed applying this cut: a track is interpolated using each top and bottom hits combination opening a window of three strips around the extrapolation in the middle chamber to accept the third hit (in the middle chamber). 3.All the track candidates passing the pre-selection are then  2 - ordered 4.The final round in the track selection consists in a loop on all the candidates starting from the best  2. The tracks is accepted if its hits were not used from previous (better) tracks. In this way a hit cannot be used to construct more than one track. EEE Analysis Meeting - 15/06/20158

9 Tracker Step 1 6 hits  5 clusters in this example Before clusterization EEE Analysis Meeting - 15/06/20159

10 Tracker Step 1 6 hits  5 clusters in this example After clusterization EEE Analysis Meeting - 15/06/201510

11 Tracker Step 2 All Top-Bottom combinations are explored and if there is a cluster matchable in the middle chamber the triplet becomes a track candidate (2 candidate in the example). EEE Analysis Meeting - 15/06/201511

12 Tracker Step 3 Track are  2 -ordered   2 1 <  2 2 1° candidate 2° candidate EEE Analysis Meeting - 15/06/201512

13 Tracker Step 4 The first candidate is accepted and its cluster are exlcuded for the other tracks. Then the next candidate is considered… In this case the 2° candidate is excluded because it looses one of its clusters in previous associations  its remaining clusters are available to form other tracks (no other candidates in this case). 1° track 2° candidate  no track 13

14 Calibration Calibration is performed in a similar way as before but the possibility to calibrate strip by strip was added (it can help in case one strip has a big noise). EEE Analysis Meeting - 15/06/201514 as before

15 Output variables: some questions Angles 1.in degrees? 2.W.r.t. the North or in local coordinates? (the angle needed for the conversion is available in the header in both cases) Currently the GPS events are not written in the event tree. Shell I re-introduce them? (the reason I removed them is that in the DQM we have systematically 2 Hz of GPS event which introduces a bias in many of our DQM plots) Other suggestions? EEE Analysis Meeting - 15/06/201515

16 Analysis The way to analysis data will be similar to the previous case (now multi-track info are available). Alternatively the framework will allow to access data in a more elegant way: EEEEventReader reader("TORI-04-2015-03-27-00001.root"); if(reader.Init()) return; EEEHeader *header = reader.GetHeader(); header->GetSchool(); header->GetYear(); // header->… Int_t n = reader.GetEntries(); // number of events EEERecoEvent *event; EEETrack *trk; for(Int_t i=0;i < n;i++){ // loop over the events event = (EEERecoEvent *) reader.GetEvent(i); event->Seconds(); event->NanoSeconds(); Int_t ntrk = event->GetNTracks(); // number of tracks in the event Int_t ngoodtrk = 0; for(Int_t j=0;j < ntrk;j++){ // loop over the tracks trk = event->GetTrack(j); if(trk->Chi2() Phi(); trk->PhiLocal(); } EEE Analysis Meeting - 15/06/201516


Download ppt "Update on the new reconstruction software F. Noferini E. Fermi center INFN CNAF EEE Analysis Meeting - 15/06/20151."

Similar presentations


Ads by Google