bfonta / hpair

Adaptation of M. Spira's HPAIR for parameter scanning

The bulk of this repository is taken from Michael Spira's HPAIR code.

My contribution was to create a wrapper script to perform parameter scans. Additionally I added a very basic I/O functionality to enable passing the names of input and out files to HPAIR's executable.

HPAIR installation

• Make sure you have the Fortran compiler gfortran installed
• Clone this repository and cd hpair
  • Alternatively, to get the original version of HPAIR, download the compressed file stored here, do a mkdir hpair, copy the downloaded file to this folder and decompress it: tar xzvf hpair.tar.gz.
• Install LHAPDF following these instructions. I personally used:

wget https://lhapdf.hepforge.org/downloads/?f=LHAPDF-6.5.4.tar.gz -O LHAPDF-6.5.4.tar.gz
tar xf LHAPDF-6.5.4.tar.gz
cd LHAPDF-6.5.4
sudo mkdir /opt/lhapdf
sudo ./configure --prefix=/opt/lhapdf
make
make install

• LHA PDF sets are available here; to install the suggested default, do (similar for other sets) sudo wget http://lhapdfsets.web.cern.ch/lhapdfsets/current/PDF4LHC15_nlo_100.tar.gz -O- | sudo tar xz -C /opt/lhapdf/share/LHAPDF (possibly adapting the paths if you installed lhapdf somewhere else)
• Compile HPAIR: make && make install
• Check the configuration stored in hpair.template.in:
  • For LO-only, set LOOP = 1
  • For SM (MODEL = 0) without new resonances, use PROCESS = 0
  • To change the trilinear coupling, one can modify C_3 = 1.0D0

Run this wrapper

You'll need to install the language go. In order to vary the "C_3" parameter (Higgs trilinear coupling) and produce a results.txt files with the resulting cross-sections, run:

go run parameter_scan.go m1 0 1 2 3

where go's native parallelism is exploited and the following happens:

• 5 HPAIR input files are stored under inputs/ (the suffix m denotes negative numbers)
• HPAIR is run over those inputs
• the standard HPAIR output files (also 5 for this example) will be stored under outputs/
• a summary of the results is stored in outputs/results.csv for easier access (in the same order as provided by the user) with the following structure:
  • 1st line: values of the born cross-section
  • 2nd line: errors of the born cross-section
  • 3rd line: values of the NLO cross-section
  • 4th line: errors of the NLO cross-section
• the values above will have different meanings depending on the hpair.template.in file