MEG Software Status Framework for MEG MC Unification

MEG Software Status Framework for MEG MC, Unification of Large. Prototype/beam test, Schedule and DC reconstruction MEG Software Group 26 Jan, 2004 1

Framework for MEG MC 26 Jan, 2004 2

Framework for MEG MC(1) • Unified framework with double option: – Tokyo or Pisa code, • Framework based on: – Fortran 77, – CERNLIB, – ZEBRA, – GEANT 3, • Distributed via CVS. 26 Jan, 2004 3

Framework for MEG MC(2) • • • Tree structure & skeleton … ready; Materials & tracking medium … merged; Event generator … stand alone & built-in; DC, TC, Magnet and B field … merged; ZEBRA structure … in progress; Scintillation photon tracking … in progress. 26 Jan, 2004 4

To be done: MEG MC • • • Merge Liq. Xe Geometry; Test of reduced number of PMT in Liq. Xe; Digitization for all the sub detectors; Full reconstruction program; Offline database… ODB? Separate database (e. g. Postgre. SQL, My. SQL, etc. )? 26 Jan, 2004 5

Large. Prototype/beam test 26 Jan, 2004 6

Large. Prototype/beam test unification(1) • MC Merge into the new framework … just finished. – Common output format for DATA/MC; – Scintillation photon tracking in Liq. Xe … Full ray tracing or geometrical tracking; – Reflection on the PMT quartz window … Fresnel or total reflection; – Scintillation light spectrum … Monochromatic, Gaussian, Basov et. al, etc. – Rayleigh scattering/absorption in Liq. Xe. 26 Jan, 2004 7

Large. Prototype/beam test unification(2) • Analyzer – Na. I calibration/gain correction … merged; – Na. I vertex/energy reconstruction … stand alone program, merging into analyzer being in progress. 26 Jan, 2004 8

To be done: Large. Prototype/beam test • • Na. I response simulation; Li. H target and Vessel simulation; phase space simulation for p 0; Offline database … ODB? Separate database (e. g. Postgre. SQL, My. SQL, etc. )? 26 Jan, 2004 9

Schedule/man power • Will be discussed in the Software meeting and reported in the review meeting. 26 Jan, 2004 10

DC reconstruction • Track Dictionary – Efficiency – Resolution • L-R ambiguity solution 26 Jan, 2004 11

The dictionary concept Several tracks with similar kinematics Several tracks with producing a single hit pattern similar kinematics producing a single hit pattern Hit pattern Single and unique string (i. e. a dictionary key) Average (over the set) 26 Jan, 2004 track parameters 12

The track dictionary is a ordered list of records: Key (hit pattern) average track parameters + rms The track dictionary exploits the “digital” response of the spectrometer 26 Jan, 2004 NO Tdrift used NO z measurements used yet 13

MC sample used to build the dictionary: • Positrons from Michel decay; • Unpolarized muons; • Generator level cuts: Ø 0. 08 < |cosθ| < 0. 35; Ø-60° < φ < 60°. 26 Jan, 2004 14

250000 generated events 12900 patterns; efficiency = 95% The population of the patterns is not uniform: 40% has 1 entry 43% has 2 ÷ 10 entries 13% has 11 ÷ 50 entries 4% more than 50 entries Number of events in a dictionary record 26 Jan, 2004 15

Momentum components for events in the dictionary LEFT Event by event distributions Track first turn has hits in at least three sectors 26 Jan, 2004 RIGHT Px / Me. V Py / Me. V Pz / Me. V Average in each Dictionary record 16

The comparison of the distributions of an average parameter in the dictionary with the actual parameter distribution shows: • Px and Py have similar shapes; • Pz • a hit pattern in the spectrometer cannot tell the sign of Pz; • the shape of the distribution of |Pz| is not well reproduced 26 Jan, 2004 |Pz| resolution of the dictionary. 17 poor

What is the dictionary “resolution” for all parameters ? Generate a sample of independent events vertex X vertex Y Px Py For tracks in the dictionary acceptance (Nsectors > 2) find the dictionary key compare; Px with <Px>(key); normalize to RMS<Px> p. MC - < pdict > σ vertex Z Pz

To be done: dictionary • Optimize stats. given by 1 – eff. ~ 10 -3 10 -4 and by looking at RMS vs. stats. (intrinsic resolution of method); • Add noise hits; • Add inefficiency of Drift Chamber; • Add drift time; • Superimpose tracks. 26 Jan, 2004 19

LEFT – RIGHT AMBIGUITY SOLUTION Digitization of the MC hit from x, y, z to: number of sector (1 -17), number of chamber (1 -2) number of wire (1 -9) D. C. A(digit) smearing of 200 mm Tdrift (const Vdrift) Z(digit) smearing of 300 mm drift circle First reconstruction step Starting from digit ID and drift time, in each sector we have 4 possible solution (4 tangent segments) 20

The assumption: track ~ circle with centre in C if PT and QT are straight segments tangent to C and intersecting in point T, then α = α´ The strategy: select the right tangents in two consecutive sectors by choosing the pair giving the minimum Da=a-a´ C ‘ Intrinsic limitations: • non uniform B implies that tracks are not exactly circles • drift distance resolution 26 Jan, 2004 a a’ P Q T

The plot shows the distribution of Da for 1000 tracks All possible combinations (23097) We need to define a cut Exact combinations on Da (3778) which allows to keep high efficiency for correct leftright choices and to reject wrong combinations Da=0. 24 rad rejects 57% of the incorrect solutions 26 Jan, 2004 22

Efficiency vs Da efficiency With Δα = 0. 24 rad we reach 90% of total efficiency in L – R solution. By definition, the total efficiency comes from two terms: -Tracks where the L – R ambigurity is solved in each sector (60%); -Tracks where the L – R ambigurity isn’t solved only in Da cut 1 sector (30%) 26 Jan, 2004 23

To be done: L-R Ambiguity solving • improve efficiency, evaluate timing; • study the effect of resolution varying with the impact parameter; • use the “calibrated” digits, (i. e. x, y. z as estimated after left-right ambiguity resolution) as starting points for F. Cei’s algorithm estimating track parameters. 26 Jan, 2004 24

A long term plan once a fit algorithm is defined • Get dictionary output – if the hit pattern corresponds to a key – when/if the resolution is appropriate for the fit (save computing time) go to the track fit else • Solve left-right ambiguity – if the track is not found in the dictionary – if the hit pattern gives ambiguous track parameters (high combinatorial calculations only when needed) go to the track fit