Project

General

Profile

Actions

Bug #351

closed

Strange behaviour of primary vertex finder

Added by Volker Friese almost 7 years ago. Updated almost 7 years ago.

Status:
Closed
Priority:
Urgent
Assignee:
Target version:
Start date:
08/28/2015
Due date:
% Done:

0%

Estimated time:
Spent time:

Description

In the current trunk, running run_reco.C after run_sim.C (deactivating vertex smearing, i.e. the true primary vertex is always at (0,0,0)) yields the follwing strange output of the primary vertex finder:

Primary Vertex Finder
Vertex coord. (0.0011955,-0.0004482,-0.0025925) cm, chi2/ndf = 0.9712384, 6 tracks used

In release JUN15, it was more reasonable:

Primary Vertex Finder
Vertex coord. (-0.0000767,0.0000688,-0.0007497) cm, chi2/ndf = 4.3339332, 206 tracks used

Why is number of tracks used for the vertex fit suddenly so low?

Moreover, if I activate vertex smearing (1 cm width of beam profile), I get the following results:

Event 0, true vertex (-0,43, 0.78, 0.0125)
Primary Vertex Finder
Vertex coord. (-4.1168895,-1.7384572,-17.3069544) cm, chi2/ndf = 168.7030764, 9 tracks used

Event 1, true vertex (0.21, -1.83, 0.011)
Primary Vertex Finder
Vertex coord. (0.4595419,-1.6903783,1.9252534) cm, chi2/ndf = 78.1232042, 8 tracks used

Something goes really wrong here.

Actions #1

Updated by Volker Friese almost 7 years ago

  • Status changed from Assigned to In Progress
  • Assignee changed from Grigory Kozlov to Volker Friese
  • Priority changed from Normal to Urgent

The behaviour in r8317 is still ok, so the problem seems to be connected to the changes in STS of r8323. I will investigate.

Actions #2

Updated by Volker Friese almost 7 years ago

  • Status changed from In Progress to Closed
Actions

Also available in: Atom PDF