TPC status report Marian Ivanov Outline Planning tool

TPC status report Marian Ivanov

Outline Planning tool Savannah bugs concerning TPC reconstrunction status (Thursday 25 June 2009) TPC calibration/alignment using Kalman Filter ( Thursday 25 June 2009)

Planning tool 2471 – New OCDB entry – time dependent gain added and used in the reconstruction – Calib/Time. Gain (To be added also to the OFFLINE OCDB) 2468 – New OCDB entry for drift velocity Array of spline fits (corrections) (data volume ~ 1 Kby) Different fits for each trigger mask (identified by strings) + one for all trigger+ one for calibration using laser Not yet in OCDB 2468 b - New OCDB Entry for space point correction (array of Transformations) (Thursday presentation) 2468 – Work in progress (special TPC HLT/OFFLINE meeting on Thursday)

Drift velocity correction Relative correction Correction using tracks crossing central electrode Data corrected for P/T Open issue – P and T in Online (HLT) different time granularity than in OFFLINE Verification of procedure using laser Not yet done Not overlap between cosmic and usable laser tests Calibration using ITS/TPC matching as Analysis. Task – Work in progress

Planning tool / Simulation 2454 – 2455 - Done 2456 – To be done (before 03. 07) 2458 – Working 2459 – Time stamp for simulated events missing – potential problems to apply calibration parameters changing during the run (e. g. drift velocity, gain)

Gain calibration

Planning tool 2460 2461 – Stored in the OCDB Procedure to be tested in P 2 (Christian Lippmann) 2462 – New RCU data format used in Das, usage in the reconstruction to be tested/commited (Jens Wiechula)

Planning tool - Reconstruction 2077 – Cuts to be tuned in the analysis 2309, 2463 – Done, to be verified in July 2464 -2467 – Ali. External. Track. Param: : Propagate doesn't take into account Br and Brfi – to be fixed by Jouri Belikov Preliminary tests with exact MC references - improvement in pt, but not in theta/z 2465 – Time dependent gain correction in OCDB, position dependend correction also implemented used (small mismatch between simulated and real data)

Planning tool. 2477 – To be tested at Point 2 Other not yet finished – (Peter Christiansen) QA checker for TPC simulation/reconstruction implemented in parallel in PWG 1 - expert mode (see Jacek Otwinoski talk) QA for the detector matching – expert mode – to be added

Online Calibration Ø Any news/updates on the procedures? Drift velocity and gain calibration (HLT) DA using raw data- Drift velocity calibration to be commited soon Ø Is everything properly integrated in the Shuttle framework? Ø For which run type would you need the GRP object (in order to be able to reconstruct the run)? Not the case for prevously mentioned entries Laser run in addition Ø Are any updates necessary in the Shuttle code running at P 2? See above

Alignment/Calibration offline procedures Ø If you did not mention this during the review, do you foresee any offline procedure especially as far as alignment is concerned? Ø Does your procedure comply with the Framework? Yes See presenetation on Thursday (Kalman filter) In case of filtering the ESD friends: Ø What is the size of the fraction you would need to store (compared to the present size/ to the ESDs size; depending on the type of collision, energy. . . )? Ø Would you need to store any other object? Pt cut, all cosmic, sample of all data (1 -5 %) Calibration components (drift, gain) Ø What are the specific requirements of your task, if any? 2 kind of tasks – light (to run on full statistisitc), heavy (to run on filtered ESD friends) – special calibration train

Alignment/Calibration offline procedures Ø What is the foreseen size of the object you will need to store? (depending on the type of collision, energy. . . ) Ø Is the memory consumption of your procedure under control? 2 implementations – one inside of the framework, combined calibration/alignment – not Analysis task (yet? ) Ø Would you need manual intervention before writing the object in the OCDB? Do you foresee to have it automatic? Gain correction calibration, PID calibration, Alignment – Thn. Sparse used Ø Is the Analysis. Task to compute the alignment/calibration constants already defined? ~1 -10 MBy Manual intervention needed for everything except of the gain and drift velocity calibration Ø What are the specific requirements of your task, if any?

Savannah Ø Is your detector concerned by any important Savannah report? Ø If yes, does this affect the quality of the physics results? Yes Pt, theta, z shift influence physical result Ø What are the current actions to solve these problems? Taking into acount Br and Brfi component of the magnetic field

Test with new propagation inside of the TPC (Jacek)

Test with new propagation inside of the TPC (Jacek)

Savannah

Performance issues Ø Is your detector affecting the overall CPU and memory (resident and virtual) consumption in the simulation (including the expert QA mode)? Ø Is your detector affecting the overall CPU and memory (resident and virtual) consumption in the reconstruction (including the expert QA mode)? No in reconstruction. Expert QA mode affected Ø If yes, what are the current actions to solve these problems? Expert mode to be done in Performance train on filtered ESD friends