Rob and Kai were unable to attend this meeting, so the agenda started with Margret's contribution.
Margret produced several different samples of simulated 2002 TB events and pushed them through the ATLAS reconstruction chain to produce AOD's and an Ntuple that is automatically produced when the AOD is generated. She compared the reconstructed energy of the EMEC and HEC in the AOD with the EMEC and HEC energies found in the Hits File (which is generated by the Geant4 simulation.) The Truth information of the Hits File gives energies that are consistent with the energy of the initial beam particle. The reconstructed energy in the AOD however appears to be too high. The worst case was for events where the beam was a 50GeV electron. The reconstructed energy for that set of events was 133 GeV. It turned out that Warren had noticed a very similar thing in his monitoring runs, where he had found a case of 150GeV pions to be reconstructed to about 285 GeV in the AOD. Margret and Warren decided to send emails to Rob about this.
Rolf showed some slides on his efforts in dealing with the HEC geometry in the 2002 TB. The "start" of the HEC1 and HEC2 modules each have a different offset in phi with respect to the EMEC. Rolf applied a correction in the Geometry file such that the shift in the modules is properly described. This correction has been successful for the HEC3 module, but still need some work for the HEC2 module. Rolf has been in contact with Sven on this problem and Sven contributed some nice plots which he produced in order to verify whether the correction was done properly.
Ian introduced us to the new group-web-page layout and took us on a little virtual tour through some of the links. We were assured that ALL of the information from the old web-page is present in the new layout. He also showed us our very own UVic TWiki pages, which he set up in addition Please go and pay a visit to the new page!
Finally some computing issues were discussed. Asok reminded us that
when we are working on
Mercury to NOT use /tmp directory to write out files (even though some
Cern web-pages might
tell you to). The /tmp directory does not have any free space!
Instead, please use /hepscratch on
Mercury. That has lots of free space.
The request to bring some databases to UVic came up again. Asok was
worried that it might be too
much of an effort to properly mirror databases. The flipside of the
coin however is that valuable time
is lost while our jobs are waiting for database connections at Cern.
Margret suggested that perhaps
we should have some copy at UVic, to at least be able to debug our
code a little faster. Such a copy
wouldn't have to be the latest up-to-date version, and it would suffice
to update it only once in
a while, say if necessary for a new release. Asok though that would
be ok and asked us to come
up with a list of databases we would like to have here at UVic.