LHAPDF
Authors:
- S. Alekhin, M.
Botje,
W. Giele, J. Pumplin, F. Olness, G. Salam and A. Vogt
D0 contact persons:
- T. Nunnemann
- A. Karchilava,
S.
Muanza
Comments:
- Parton Density
Functions
library (proton, pion, photon, nucleus) including
their
experimental uncertainties so as to propagate those properly in
observables
systematic uncertainties determination
- Needs an
external
evolution package:
- Theoretical
prescription
for proper usage of PDFs:
Use PDFs of same order (in
perturbative
development) as your matrix element
Use PDFs with same renormalization scheme as
your
matrix element (higher orders)
Use alpha_S calculated at same order and in
the
same scheme
Beware of "stepping out" of the relevant
kinematical
values (x,Q^2) of your PDFs - Programming Language: Fortran
- Current D0
version:
v3 (2005)
LO proton PDFs: CTEQ6LL
(future D0 default), ...
NLO proton PDFs: CTEQ61 central
(future D0
default),
...
Updating the PDFsets:
- The following shell
script enables to easily fetch the tarball containg the up-to-date
PDFsets
- A soft link in the MC generators executables directory is
necessary for them to use LHAPDF v3
Proposal for new codes:
- Currently we're using the "PDFCODE"
variable
in the cardfiles (see
most utilized ones). The mapping between the PDF codes and the
corresponding
PDF names are made in production tools like mc_runjob and some shell
script
to run the generator executables.
- We're now permanently moving from the PDFLIB to LHAPDF v3. Most
of
the p/pbar PDF we used in PDFLIB are available in LHAPDF v3. However their codes have completely changed!
- To avoid any confusion among developpers and especially for the
MC
users I propose to
- keep on refering to the PDFLIB codes as "PDFCODE"
- to refer to the LHAPDF codes as "LHAPDFCODE"
- to implement the "LHAPDFCODE" in
the
cardfiles
- to request a new version of mc_runjob that adopts these new
codes
- A correspondence table can be found here
Known Bugs and Caveats:
Links:
Last update: SM 24 May 05