Running the Tracker Data Quality Monitoring ORCA code

Running the Tracker Data Quality Monitoring ORCA code

This uses a COBRA-driven event loop.
It can be run on real (FED9U data format) or simulated data.
It books and fills the histograms via the DaqPrototype interface to ROOT in COBRA.

Requires you to have downloaded all the software using the getOrcaNew and getOrcaFed9U scripts from here.

It can run either the standard Tracker Data Quality Monitoring code (formally TT6 analysis) or the simple test-beam analysis example. These are descibed on the test-beam software web page in under `TT6 Analysis' and `General.cpp' respectively. The standard monitoring code is now in sub-system TrackerMonitoring. Documentation on the code structure there is in TrackerMonitoring/doc/

The analysis programs are in TrackerMonitoring/FrameworkSiStripMon/test/. The important files are:

MonitorData.cpp (standard montoring analysis on real data)
MonitorMC.cpp (standard monitoring analysis on MC - doesn't yet work as have no DetUnit-Rod-Layer map for MC, which is needed to structure output histograms)
MyAnalysisData.cpp (simple example analysis on real data)
MyAnalysisMC.cpp (simple example analysis on MC)

None of these files contain any code. They are just linked to different libraries in the BuildFile and these create the analysis object (via PkBuilder) which then observe events dispatched via COBRA.

For analysing real data or MC, you need to use the .orcarc cards in orcarc_cobradriven_data or orcarc_cobradriven_mc respectively. Note that you also need to source the script setGeom.csh ince DaqPrototype still expects to know where the CMSIM geometry is (in case it needs the B-field map).

Warnings:

The BuildFiles use the libraries libtestDataHandlingInterface, libtestTkFilterUnit and libTkFilterUnit. These are not compiled as part of the official ORCA release, so you must compile them.
When running MonitorData.cpp, the run and event number are no longer available. Neither are various things like the PLL delay settings. These all used to be taken from the old FEDPMC data ROOT reader, which is no longer used.
Due to a bug in ROOT for version less than 4.02, profile histograms appear empty by default if they contain only one entry in each bin. You may need to plot them with the draw option "HIST" to see anything.