Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] aeeo_v3_0.tar.gz(1067 Kbytes)
Manuscript Title: Tools for NLO automation: extension of the golem95C integral library
Authors: J. Ph. Guillet, G. Heinrich, J. F. von Soden-Fraunhofen
Program title: golem95-1.3.0
Catalogue identifier: AEEO_v3_0
Distribution format: tar.gz
Journal reference: Comput. Phys. Commun. 185(2014)1828
Programming language: Fortran95.
Computer: Any computer with a Fortran95 compiler.
Operating system: Linux, Unix.
RAM: RAM used per integral/form factor is insignificant
Keywords: NLO computations, One-loop diagrams, Tensor integrals, Higher spin representations, Effective theories.
PACS: 12.38.Bx.
Classification: 4.4, 11.1.

External routines: Some finite scalar integrals are called from OneLOop [1,2], the option to call them from LoopTools [3,4] is also implemented.

Does the new version supersede the previous version?: Yes

Nature of problem:
Evaluation of one-loop multi-leg integrals occurring in the calculation of next-to-leading order corrections to scattering amplitudes in particle physics. In the presence of particles with spin two in the loop, or effective vertices, or certain gauges, tensor integrals where the rank exceeds the number of propagators N are required.

Solution method:
Extension of the reduction algorithm to rank r ≤ 10 for N ≤ 4 and rN + 1 for N ≥ 5, which is sufficient for most applications in Beyond the Standard Model Physics.

Reasons for new version:
The previous version was restricted to tensor ranks less or equal to the number of propagators.

Summary of revisions:
Tensor ranks > N are supported, an alternative reduction method for the case of infrared divergent triangles is implemented, numerical stability for the case of small mass differences has been improved.

Running time:
Depends on the nature of the problem. A single call to a rank 6 five-point form factor at a randomly chosen kinematic point, using real masses, takes 10-3 seconds on an Intel Core 4 i7-3770 CPU with a 3.4GHz processor.