(present: AA, TI, NK, RK, ML, TS)

Status Reports
Tayfun met with Margret regarding shower shape studies. He plans to produce code that will return the radiation length up to a given point in the detector. This involves translating fortran code used by Margret. For the beam chamber information, the plan is to skip events for which the beam chamber information is the same as in the previous one (initialization problem in athena) and to only consider events for which the beam chamber fit passes through the trigger finger counters (bad fits partially caused by run dependent noisy wires).

Tamara is adapting her pion energy reconstruction code to use the new OccupancyAlg recently committed by Michel. She has also looked at Margret's occupancy and channel list making code and she found it useful.

Naoko is continuing her clustering studies, first for electrons. Naoko has also confirmed that the energy weigted y reconstruction of the emec layers agree with the Monte Carlo up to a global offset. This will be included in the global geometry file by Michel soon.

New systemAlg's
Michel presented two of the systemAlgs he recently produced and committed.
NoiseAlg computes and allows access to digital filtering noise (nA) (and its error) for each (connected) channels. Two examples (UserAlg.*) are on the TBRootAna example web page. Michel also presented some figures to illustrate the method followed by NoiseAlg, on three different data runs. The figures can be accessed here:
run 12253 (119 GeV e+),ps file
run 12366 (180 GeV pi+),ps file
run 13188 (150 GeV mu+),ps file
For each file, you have the following figures:
First page for the hec:
The noise (nA) vs the pedrms (nA) for the channels judged to contain no signal (hence good for noise estimate);
The gaussian fit (+/- 2 sigma) mean (nA) vs the histogram mean (nA) for the same selected channels;
The histogram mean (nA) vs the ratio noise/pedrms for the same selected channels;
The ratio noise/pedrms vs eta for the same selected channels;
The mean per layer per eta bin of the ratio noise/pedrms vs eta which is then used to interpolate to the noise for the channels that have signal (and hence that cannot be used directly to estimate their noise) using their pedrms value;
The final result noise (nA) ve eta for ALL channels.
The second page has the same plots for the emec.
The other pages show examples of signal distribution in channels. The vertical axis label indicates the NoiseAlg status of this channel: 0 means that the channel was judged to have no signal and was therefore used to estimate its noise, while >0 means that the noise value for that channel had to be obtained by interpolation.
OccupancyAlg computes and allows access to the occupancy (and its error) of each channel for a given run. Two examples (UserAlg.*) are on the TBRootAna example web page. One of these examples uses OccupancyAlg to obtain a list of channels that have an occupancy (obtained with a 2 sigma cut) greater than 10%. Maps of the occupancy of these channels for run 12366 (180 GeV pi+) have been shown and can be accessed here in two ps files, one for the hec and one for the emec. Only physics triggers and good global cubic timing events are retained by OccupancyAlg. On these figures you can see that some emec channels have pathological behavior: channels with large but otherwise gaussian noise do not populate the high occupancy channels.

TBRootAna update The TBRootAna release is awaiting the finalization of the neighbor methods, which are being (greatly) improved. Apart from NoiseAlg and OccupancyAlg, the systemAlg RunStatisticsAlg is also available.

Athena update Naoko reported that athena release 6.5.0 is due this week. Naoko also reported that the initialization problem in the athena beam chamber code has been fixed in 6.5.0. Naoko has developed code to isolate noisy beam chamber wires for a given run. This knowledge can then be used to reanalyse the data and produce an ntuple with better beam chamber data. As this is a lot of work and since TBRootAna work can proceed with the current ntuples, Richard will think about what is the best manpower strategy to tackle this problem, if required.