Flux usage in bggn/bggn_upd

[igjaegle]

The different bggen generators simulate a fixed flux that is hard-coded in the generator card. Therefore, to obtain the correct flux for each run, a new card must be used. Could cobrems be implemented in the various bggen generators?

[aaust]

gluex_MCWrapper creates the cards at run time according to the flux stored in ccdb. Does this request go beyond this feature?

[igjaegle]

MCwrapper calculates the number of events to be generated for each simulated run number, but it does not rewrite the generator card with the corresponding untagged flux distribution for that run.

The ideal would be that the generator used the untagged flux according to the data-base as all the other generators.

[jrstevenjlab]

Right, this goes beyond an MCWrapper feature because the bggen generators (written in Fortran) don't use the BeamProperties class to which is common to all our other generators (written in C++) for choosing between flux options (cobrems calculation, CCDB flux or input file).

I think Lubomir does have a mechanism for the J/psi generator to use a histogram of the flux as an input for generating those samples. In that case you can control the total flux distribution, but don't have the information for the generator on a run-by-run basis

[igjaegle]

This mechanism consists of hard-coding in the generator card the distribution. So one needs to produce one card per run. I am doing that easily on the farm ... but this solution is not clean for OSG. Hence the needs to implement in the fortran code the Beamproperies.