Updates for lxplus

[mvicenzi]

This PR addresses the items in #42.

• The dependencies were updated to remove GENIE. The setup script for lxplus (lxplus_setup.sh) relies on LGC software stacks in cvmfs and a locally installed HEP_HPC package. It also sets the environment for xroot-based access to /eos/user/ (important for data transfer to/from job nodes).
• Input gst GENIE files have been moved to /eos/user/m/mvicenzi/genie, which I made available to other users.
• README.md has been updated with new setup instructions.
• Macros have been updated to use latest parameters names and default geometry configurations. The set of single macros for neutrino events and single particles are ready-to-use. I moved everything else to a subdirectory (legacy).
• I've updated vis.mac with visualization drivers available on lxplus. It takes quite a long time to load the geometry (at least for me), so there might be something else to tweak.
• I reworked the job submission scripts for the lxbatch system (which is based on HTCondor) in GridUtils/lxplus. I put a single-job example (described in the README.md as well) and also a subdirectory for multiple-job submission of neutrino events (any flavor). All the parameters for this type of production can be set directly at the top of nu_job.sh.

The package is now fully portable to lxplus.cern.ch with all its functionalities! Please test what you can to make sure the instructions are clear and that there no permission issues to access certain areas or files. Thanks!

[mvicenzi]

Recent updates:

• I've added the +JobFlavour parameter in the .sub macro for the submission of neutrino events as the default value was "espresso" which meant a max of 20 min wall time for the execution (clearly not enough sometimes). I put "microcentury" which corresponds to a max wall time of 1 hour.
• I've updated the job numbering so that log files, output files and mac files share the same numbering from 0 to n_jobs. This should prevent confusion when debugging with logs.