[Halld-physics] Analysis Launch Comments

[Sean Dobbs]

Dear all,

Since we now have a large set of the GlueX-I data reconstructed, there
are some general comments that have been touched upon at recent
Analysis & Production meetings which should be kept in mind, in order
to ensure efficient use of CPU & disk resources.

- It is generally considered good practice to test a new reaction out
on a single run's worth of data or on the 2017 subset of data before
requesting a run over the full GlueX-I data.  This is particularly
true for high multiplicity final states or those with missing
particles, which often result in very large ROOT trees.

- It is suggested to not request channels containing neutrons or
antineutrons until particle ID criteria for these particles have been
established.  The default cuts are extremely minimal and lead to
excessively large output files.

- For specialized studies, such as those for tracking or particle ID
studies, analyzers are encouraged to develop their own plugins in
which they can place additional event selections to reduce the tree
size, and/or modify the output format.

Cheers,
Sean