Allow manipulation of integrals and decay partial widths / BRs

[kilian]

Currently, for determining branching fractions and for determining event numbers in simulation, WHIZARD uses the integrals it has calculated itself. The user should be able to modify the results, so he can manually insert K factors or measured cross sections / branching fractions while keeping the benefit of WHIZARD's phase space sampling. In particular, he could make use of the decay widths stored in the SLHA record.

(The 'weight' expression can be abused for this purpose, but it would allow only for relative adjustments, not for setting the value directly.)

[kilian]

Treating the full decay width as an independent parameters would also allow to weight process cross sections by branching ratios, thus approximating exclusive matrix elements by factorization.

[kilian]

Also implement a method to specify numeric codes for processes, using a similar syntax.

[Juergen Reuter]

As this Wolfgangs next task, I rank it higher.

[kilian]

This is done in r2475.

The result variables integral(proc) (also n_calls, error etc., but this is probably useless) can be reset by the user once they become available due to a previous integrate command. The simulate and unstable commands will use the redefined values. This should cover all use cases.

Furthermore, each process that is defined by a process declaration acquires an integer variable num_id(proc). The value is undefined initially, but the user can give it a value, locally or globally. Reading and writing to HepMC or HEPEUP format, the user-defined IDs are activated by setting ?use_num_id. This way, event files written by external generators can be rescanned, and WHIZARD's event files can be given arbitrary user-defined process IDs.

By default, the numeric ID is just the position of the process in the simulate or rescan command.