[d2n-analysis-talk] BigBite Farm Replay: MWDC variables

[Diana Parno]

Hello all,

I ran a test of the farm replays with the modified .odef file I sent  
to the list last night. Much to my surprise, the result was an  
additional factor-of-3 reduction in file size (on top of the factor- 
of-4.5 reduction from cutting out events with no tracks). My fully  
replayed sample run, 2024, now takes up only 2.6 GB of disk space in  
root files. (Before the above changes, it occupied 30.7 GB.)

I've now investigated this result, and I believe that the main source  
of file size reduction is that the BB.mwdc.* block is no longer being  
written to the root files. This block represents 628 branches, and the  
majority of these are arrays, so they do take up quite a lot of room.

This level of reduction in file size would go a very long way in  
solving our immediate space problems. Before moving forward and  
submitting these jobs to the farm, though, I want to make sure this is  
a sensible step. Are the MWDC calibrations currently in a finished  
state, or do we expect to need these MWDC files for a wide range of  
BigBite runs? If the latter, are there particular variables we expect  
to need (so that we can strip out the others), or do we require the  
whole kit and caboodle? Matt, I think you're the MWDC expert, so I'm  
especially hoping you can weigh in.

Best,
Diana