CompHEP

CompHEP
Name	CompHEP
Title	CompHEP
Developer	Novosibirsk State University; Joint Institute for Nuclear Research; CERN contributors
Released	1990s
Latest release version	(varies)
Programming language	C, Fortran
Operating system	Unix, Linux, Windows (via ports)
Genre	High-energy physics software
License	Open-source (varies by distribution)

CompHEP is a symbolic and numerical software package for calculation of scattering amplitudes and cross sections in high-energy particle physics. It provides automatic generation of Feynman diagrams, algebraic simplification, and numerical phase-space integration for predictions used in collider experiments. The project has connections to research groups at Novosibirsk State University, Joint Institute for Nuclear Research, and CERN, and contributes to workflows in collaborations such as ATLAS, CMS, and LHCb.

History

CompHEP originated in the early 1990s amid efforts at Novosibirsk State University and Joint Institute for Nuclear Research to automate perturbative calculations for processes studied at LEP and planned at LHC. Early development paralleled contemporaneous projects like GRACE, MadGraph, and FeynArts while interacting with institutions such as IHEP (Protvino), DESY, and SLAC. Over successive releases the codebase incorporated algorithms inspired by work at ITEP, INR RAS, and contributions from groups collaborating with CERN and FNAL. Workshops at venues including Les Houches and Sakharov Conference fostered interoperability efforts with standards promoted by HEPData and groups involved with Monte Carlo validation. Throughout its history CompHEP was influenced by theoretical advances from authors affiliated with Moscow State University and experimental needs from Tevatron and LEP2 analyses.

Features and Capabilities

CompHEP automates symbolic generation of tree-level Feynman diagrams for user-specified processes, producing squared matrix elements that can be exported to numerical codes. It supports models defined in formats compatible with model-building efforts from SLAC National Accelerator Laboratory groups, with capabilities reflecting interactions studied at CERN experiments and theoretical frameworks from Institute for Theoretical and Experimental Physics. The package provides phase-space integration using Monte Carlo techniques used in PYTHIA and HERWIG workflows, event generation suitable for detector simulation chains used by ATLAS and CMS, and interfaces for parton distribution functions from CTEQ, MSTW, and NNPDF efforts. Users can define custom Lagrangians influenced by model-construction literature from John Ellis, Howard Georgi, and Steven Weinberg-style frameworks, and export processes compatible with analysis toolchains at FNAL and DESY.

Architecture and Implementation

CompHEP combines symbolic algebra routines implemented in C and Fortran with numerical integrators and event generators. Its architecture separates model definition, diagram generation, algebraic simplification, and numerical modules—an approach reminiscent of designs at MadGraph and FormCalc projects. The code interfaces with libraries used at CERN such as ROOT and with PDF access methods common in LHAPDF distributions maintained by groups at University of Oxford and NIKHEF. Build systems have been adapted to work on clusters at CERN and supercomputing facilities at JINR, with portability to environments used by Brookhaven National Laboratory and Fermilab.

Usage and Workflow

Typical workflows begin with model selection or creation reflecting theories pursued by researchers at Institute for Nuclear Research (Moscow), followed by process specification informed by experimental analyses at ATLAS or CMS. Users generate diagrams, inspect algebraic expressions, and run numerical integration to obtain cross sections and distributions comparable to results from Tevatron and LEP publications. Output events integrate into detector simulation chains like those used in GEANT4 and reconstruction workflows developed for LHCb. Collaborations often combine CompHEP outputs with parton showering in PYTHIA and higher-order corrections from programs associated with NNLO and NLO projects driven by theory groups at CERN and DESY.

Validation and Performance

Validation of CompHEP results has relied on comparisons with analytic calculations by authors from Moscow State University and cross-checks against numerical outputs from MadGraph, CalcHEP, and GRACE. Benchmarks have been performed using datasets and test cases inspired by studies at LEP and Tevatron analyses, with computational performance profiled on clusters at CERN and national facilities such as JINR and INRIA-associated resources. Performance scaling depends on model complexity and phase-space dimensionality; optimization strategies mirror those used in Sherpa and Whizard to reduce integration variance and CPU time.

Applications and Impact

CompHEP has been used in phenomenological studies by researchers associated with IHEP (Protvino), Moscow State University, and experimental collaborations at CERN and FNAL, contributing to predictions for processes relevant to searches for physics beyond the Standard Model pursued by ATLAS and CMS. It informed studies in electroweak physics, heavy-flavor production, and new-physics signatures considered in programmatic reviews at Les Houches and influenced pedagogy at institutions like Novosibirsk State University and Moscow Institute of Physics and Technology. Results from CompHEP-supported analyses have appeared alongside outputs from MadGraph and Sherpa in phenomenology literature authored by scientists affiliated with IHEP, DESY, and SLAC.

Related Software and Integration

CompHEP shares functionality and integration pathways with programs including CalcHEP, MadGraph, FeynArts, FormCalc, Sherpa, Whizard, PYTHIA, and HERWIG. It interoperates with analysis and data formats used by ROOT, LHAPDF, and detector simulation toolkits such as GEANT4. Collaborative software ecosystems linking to CompHEP have been discussed at workshops hosted by Les Houches, Sakharov Conference, and institutions including CERN and JINR.

Category:High-energy physics software