[Halld-offline] proposal for data challenge details

[Mark M. Ito]

We discussed some details of the data challenge configuration at the 
last meeting. See

 
https://halldweb1.jlab.org/wiki/index.php/GlueX_Data_Challenge_Meeting,_November_19,_2012#Data_Challenge_Configuration 


We agreed that I would send out a proposal for some of the specific 
choices. Here they are including the choices we made at the meeting.

= 1. run numbers =

All runs will be run 9000.

= 2. random number seeds =

The trick here is to insure that the we do not repeat any random 
sequences in any of the jobs across all sites. The proposal is

a) Use the file number to seed bggen (See definition of file number below.)
b) Do not explicitly seed hdgeant.
c) Do not explicity seed mcsmear.

= 3. file naming convention =

a) for the rest format data: rest_<filenumber>.hddm
b) for the monitoring histograms: hdroot_<filenumber>.root

where <filenumber> is an incrementing six digit integer, padded on the 
left with zeros, with an initial value specific to the site:

grid: 000001
cmu: 100001
jlab: 200001

Jobs and filenumbers are in one-to-one correspondence.

= 4. events per file =

200,000

chosen to take about one day at JLab

= 5. CKOV in control.in =

CKOV should not appear in control.in.

-- 
Mark M. Ito
Jefferson Lab
marki at jlab.org
(757)269-5295