Algorithm works both on real data and MC.
1. Algorithm

The Outer tracker segments are prolongated to the Z position of cluster and are compared with cluster position. The momentum of the OTR segment is assigned to the matched pair as RSEG->PF. In case it OTR momentum is not available it is estimated by target constrain. No comparison with ecal energy is performed.

2. Selection of objects to be matched
1. OTR : taken from RSEG ArteTable
   1. rseg->fit = 0
   2. (rseg->cmp & Rsegc::bitsign) == 0
   3. rseg->cmp == Rsegc::patt || rseg->cmp == magt
   4. -30 < rseg->ze < 1500
   5. the end point of RSEG is considered for match
2. ECAL : taken from RCCL ArteTable
   1. rccl->flag > 0

3. Match Parameters

DEFAULT
OCmatchConst

• match based on { X, Y \} variables;
• Z of plane of match = Z of cluster
• abs( delta_X ) < 4 cm; abs( delta_Y ) < 20 cm;
  
• no $\chi^2$ cut;
• no check on combination with given object to be matched;

optimized with ARTE-03-05 on real data

SOFT
- OCmatchSoftConst
• match based on { X, Y \} variables;
• Z of plane of match = Z of cluster
• abs( delta_X ) < 20 cm;
  abs( delta_Y ) < 20 cm;
  
• no $\chi^2$ cut is used;
• no check on combination with given object to be matched
optimized with ARTE-03-05 on real data

4. Call Sequence
-

RECON/MATCH marple OTR ECAL
from user.kumac, if standard reconstruction for real data is called
or
void matchotrecal_();
from usevnt_() if no reconstruction is called.

5. Recomendation
The algorithm was tested on the real data december'99 with ARTE-03-05-r1. The programs used for subdetector parts were RANGER and CARE. Since for data december'99 -april'00 the TY resolution is not very good it could be replaced by Y_otr / Z_otr.


In the standard reconstruction the different match procedure provided by CARE is supposed to be used. It uses internal care structure and allows to extract MIPS. However, to check correlation of OTR-ECAL or in case the alighment is not proper the marple algorithm is more efficient for non-MIP clusters.