Skip to content

Latest commit

 

History

History
58 lines (46 loc) · 2.05 KB

File metadata and controls

58 lines (46 loc) · 2.05 KB

MadGraph7

MadGraph7 is the next major release of the MadGraph event generator. Its new features include vectorization and GPU support for matrix element evaluation and phase-space sampling, machine-learning-accelerated phase-space integration and sampling through MadNIS, and more accurate handling of spin correlations in MadSpin.

Alpha release

This is an alpha release, meant for testing and feedback, not for production. The MadGraph7 workflow currently covers leading-order event generation. The LO workflow from MG5_aMC@NLO is still reachable through the output madevent mode. The NLO workflow is also available (with only minor updates). AmpliCol mode is not included in this release.

If you need a stable release, use MadGraph5_aMC@NLO (also on Launchpad).

Installation

Download the latest release tarball from the releases page, unpack it and start the interface:

tar xzf MG7_v0.2.0.tar.gz
cd MG7_v0_2_0
./bin/madgraph

Python 3.12 or higher and a C++ compiler are required. The madspace library is installed automatically the first time you launch a run. To build it yourself instead, run install madspace from the MadGraph7 prompt.

First steps

Type tutorial at the MadGraph7 prompt for a guided walkthrough of the commands.

To start your first process, you can enter the following commands:

generate p p > t t~
output my_first_run
launch

MadBoard is a separate package providing a modern web interface to MadGraph, for setting up processes, following runs from the browser and inspecting their results. You can install it with

pip install madboard

and launch it by running the madboard command within your MadGraph7 directory.