GitLab

Original commit message:

Add namespace specifications for instantiations

Add namespace to explicit class template instantiations, as otherwise clang
sees the omission as an error, if using

    CXXFLAGS="-std=c++11 -stdlib=libc++".

This enables compilation with clang linking to Rivet 2.5.0, which requires C++11.

For the same reason, an implicit cast has been made explicit and some weird
rvalue construction has been replaced with code that is readable and compiles.

This only affects AlphaS reweighting with parton shower emissions.

This fixes compilation when --with-lhapdf is not defined

Specifically, we do not generate new xp values if we cannot reuse them from the
nominal run. Instead, we do not calculate collinear counterterms for the
reweighting in this case. This ensures statistical equivalence to prior
revisions.

More specifically, the re-generation of the random numbers used to calculate
the xp values for the collinear countertems, whenever a scale variation is
done, has been removed. Instead, the (anyway stored) xp values are used.

svn merge -c26207 ^trunk/SHERPA: Give a more meaningful name to our default PDF (used by Sherpa::PDFInfo()).

Original commit message: "Update reweighting citation"

  c28567: allow custom NNPDF grid path
  c27742 + c28565: allow custom DECAYPATH

merge trunk c28357: Make Dir2Db and Db2Dir work with special characters in filenames (e.g. square brackets).

svn merge -r 27912:r28290 ^/branches/beyond_fixed-order_reweighting .
svn merge -cr28373 ^/branches/beyond_fixed-order_reweighting .
svn merge -r 28530:28531 ^/branches/beyond_fixed-order_reweighting .

Single HepMC2_Interface.C chunk regarding the output of RDA AlphaS values from:
svn merge -cr28454 ^/branches/beyond_fixed-order_reweighting .

- Revert any changes to cut-off behaviour of CF_QCD/Running_AlphaS to retain
  nominal physics of rel-2-2-0
- Disable shower emission reweighting per default
- Disable reweighting warnings output per default

svn merge -c28149 ^/trunk/SHERPA
svn merge -c28202 ^/trunk/SHERPA
(proper treatment of psi itmin/max by node in mpi mode)

made readin check in histogram precision dependent

svn merge -r28139:28146 ^/branches/rel-2-2-0_anl

for i in 27849 27853 27862 27871 27872 27874 27878 27880 27886 27889 27891 27897 27903 27908 27911 27922; do svn merge -c trunk/SHERPA .; done
svn merge -r27924:27934 trunk/SHERPA .
svn merge -r27946:27948 trunk/SHERPA .

check (x,Q2) whether bad before asking LHAPDF, check q2 before calcing alphaS