Monte Carlo Embedding in L00

There are two ways to use it. One way is to read in a cdfsoftware file with DHInput and read in a to-be-merged file with the EventMerger module. The second is to use a MC generator to generate events and then merge in a second file with EventMerger.

This has been superseded by new methods in 4.10.x; ask Jason Nielsen for details

Some notes:

• In either case the merging occurs at the raw data bank level so any clustering and pattern recognition will have to be redone after merging. The characteristics of the event (run number, event number, etc.) are those of the DHInput data stream. Only the specified banks of the EventMerger input file are merged in. This means calibrations appropriate to the primary stream are applied. The EventMerger will merge COT and calorimetry banks as well. I do not know the status of COTD/COTQ merging as I never tried tracking a merged event.
• Note that after clustering, one can tell SiClusteringModule to write out the SIXD banks after correction. This is what I used previously. But you have no control over the descriptor of this bank ("CORRECTED") so if your input data has a corrected SIXD bank you will run into confusion unless you drop it at input.
• The WriteBanks method for StorableRun2SiStripSet has a bug when writing mixed data/MC SIXD/ISLD banks which prevents correct MISL/MSVX banks from being made in this case. The user can still by hand figure out what is going on from using the unmerged banks if they are available. Still a patch is needed to write the mixed banks (not a fix just inhibits creation of MISL/MSVX). The new version of the TrackingObjects/src/ StorableRun2SiStripSet.cc should replace the 4.10.0 file.

A working version of the EventMergeMods package is tagged into 4.10.0. I have not tested the example ntupling code since 4.6.2, but it compiles and links.

An example ntupling module and executable code for 4.10.0 is found here This package creates two executables, l00spytest and l00spytest_gen, the latter of which is based on cdfSim and the former assumes reading in of raw data at the input. To use this, make a new release in 4.10.0, addpkg TrackingObjects and replace the StorableRun2SiStripSet.cc file as per above, and untar the package linked above into the release.

This ntuple module will loop over all L00 sensors looking for MC hits. If one exists it will record the GEANT information for each MC hit (position, entrance angles, energy, delta rays), the sensor strip data from primary, secondary, raw merged and corrected ISLD banks (charge, noise, bad, etc.), and the clustering information for the sensor. Again, no tracking is done so no track information is recorded.

An example tcl file for running the generator version (which includes all that is needed for the simpler read in data) is here I don't know if this tcl will work with cdfSim as 4.10.0 has it; this worked for 4.6.2 but that should only affect the generator settings which I assume the average user will mess with anyway. This example created a single FAKE_EVENT muon with physical deposition model.

New Jason Nielsen has figured out how to run the tcl in 4.10.4 and gives needed changes in this email.

Please contact me if there are problems so I can update this.