Evaluate matrix element / diff. cross section for given phase space points

[ohl]

This is necessary for point-by-point regression tests and/or comparisons.

• new sindarin construct

     value = matrix_element (<filename>)
     value = cross_secion (<filename>)
  

where the file is either derived from a well known event format or looks like

num_points 42
1 num_in 2
1 num_out 3
1 1 "<particle_name1>" <PDG1> <E1> <p1x> <p1y> <p1z>
1 2 "<particle_name2>" <PDG2> <E2> <p2x> <p2y> <p2z>
...
1 5 "<particle_name2>" <PDG3> <E3> <p3x> <p3y> <p3z>
1 num_in 2
1 num_out 4
...
...

• it's proabably useful to be able to write an output file containing the momenta and the matrix elements / cross sections.

[ohl]

Phone conversation Wk/TO: use HepMC as input format (color and helicity stored in the flow list).

[kilian]

As an intermediate step, the squared matrix element can now be retrieved by the user in the analysis expression: variable event_sqme (besides event_weight). Input/output beyond raw event files is not yet implemented.

I won't finish that before 2.0.1, but will stay top priority for 2.0.2.

[kilian]

Update: This is finished now, along with a major rewrite of the integration/simulation modules, and with further useful features. I'll check it in and merge the branch with the trunk once I got ssh connection again.

[kilian]

Merging done in r2468.

There is one change in the previous behavior: the process index written to file is now the index in the simulation command, e.g., in simulate (foo, bar, xyz) the process bar has index 2. This holds both for writing and reading (rescan). This way, the index is unique and easily determinable.

To allow for reading files written by other programs with their own process indices, I will implement a way to override this default behavior by an explictly specified index. This goes along with #264.