Hi Flemming + > Modified the Clear() method in BrTPCClusterTable to not delete the > Tlist keeping the clusters. It happened (with a simulation) the that > code went into looking for table when in fact no clusters had been > added. Why has this not been found before is a mystery. Kept some > assert() which was inserted to find problem. A few cases of not > finding a prefered track in BrTrackFinderBase appeared. this happened > when chisq>1,000,000 (cases where dx==dy==0 !!!) which I thought was > taken care of. > This is indeed very strange. The first problem have been fixed before (I think actually twice). The second problem I have seen for dy==0 after the last month of updates. ******************************************************************* NB! Everyone, please check carefully before updating brat that problems with Warnings are not reintroduced !!! ******************************************************************* I took a look at the new brat. Row : 11 Status : 5 Pos : (4.04555, 11.6272) Dpos : (0.000562591, 8.36408e-05) Warning in <BrLocalTrack::Fit>: Hit Width dy == 0 Row : 11 Status : 5 Pos : (16.5384, 11.5656) Dpos : (0.000500203, 4.79294e-05) Warning in <BrLocalTrack::Fit>: Hit Width dy == 0 Row : 11 Status : 4 Pos : (-2.73774, 11.5189) Dpos : (0.000292849, 9.70271e-05) Warning in <BrLocalTrack::Fit>: Hit Width dy == 0 ....and so on... It seems that the error only occurs for deconvoluted clusters (status = 4, status = 5). I tried to look for what was wrong but did not succeed in identifying the problem. Anyone who has updated recently should take a look at BrTPCClusterFinder.cxx (BrTPCHitClusterFinder.cxx seems to be untouched). Cheers Peter -- :-) --------------------------------- )-: |Peter H L Christiansen aka PAN @ NBI | |EMAIL : pchristi@nbi.dk | |OFFICE : Tb1 @ NBI | |PHONE : 353 25269 | |SNAIL : Hans Tavsensgade 35, 4th | |PHONE : 35 349336 | :-D --------------------------------- \-:
This archive was generated by hypermail 2b29 : Wed May 09 2001 - 09:58:03 EDT