[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

mucap realtime analysis



For the realtime analysis, is it really necessary to save the
tree file?  I mean, do you intend to run over the same file
more than once?  I would think not, and in that case I would
suggest the following:
1) the tree file ("-T ...") can probably be set to /dev/null.
   If it improves performance significantly, the tree creation
   and filling (etc.) could be commented out in
   src/uiuc/EventTree/TEventTreeAnalysis.cpp.  The EventTree
   subanalysis should not be disabled, though, because we
   still want the TMucapEvent object to be filled for every
   data block.
2) the tree analysis code could be called as a submodule.
   This would be a matter of creating a MIAS subanalysis
   to send the TMucapEvent object to each of the tree analysis
   submodules (as in mu/src/uiuc/macros/EventTree/src/main.cpp).
   Be careful which file these histograms are associated with:
   we want them in the main output file "-o ...", not the
   /dev/null.  It may be better to just comment out the creation
   of the "-T" file altogether if we don't care to make trees
   in this mode.

Does the above agree with the desired function of realtime analysis:
1 raw data file --> 1 ROOT histogram file?

Steven