automatically set mc_group_subproc to False if vector_size>1

[oliviermattelaer]

• [x] automatically set mc_group_subproc to False if vector_size>1
• [x] check that vector_size=0 is working
• [x] check that fixed_scale is working
• [x] check that dynamical_scale=3 is working
• [x] check that dynamical_scale=-1 is working
• [x] check that cross-section is consistent for g g > t t~ (independently of the vector_size)
• [x] check that cross-section is consistent for q q~ > t t~ (independently of the vector_size)
• [x] check that cross-section is consistent for p p > t t~ (independently of the vector_size)

[oliviermattelaer]

q q~ > t t~ fails to work correctly with CKKW scale. But the problem is NOT in setclscales() (as I thought) even if I bypass that function call the cross-section is still much too large

same issue for fixed scaled.

If i bypass the wgt (set to 1d0) and pdf factor (set to 1d0), I do have an agreement with standard madgraph. restoring the jacobian leads to very close integral while integrating only the pdf leads to result different by a factor 100...

not related to the fact that they are multiple flavour since this is also problematic for u u~ > t t~

• issue related to the pdf call (changed due to electron beam and creating a merge conflict with the vectorization that was handle in the wrong way:
• !LP=SIGN(1,LPP(IB(%d)))

[oliviermattelaer]

fix in 64bc9caa020e23a708fc4aa7e3b2b00306bed2ba