Anomaly detection
As this course will require many tools, the most convenient is to install Docker and a virtual machine with everything already pre-installed.
Instructions are here: https://github.com/valsdav/vbscan-school
Please note that the VM is about 1.2GB. Make sure you have sufficient space on your hard drive.
From within the VM, clone the tutorial repository
git clone https://github.com/UniMiBAnalyses/ML4Anomalies
git checkout -t origin/VBS_School_August_2021
the data Ntuples are at
https://drive.google.com/folderview?id=1u2WdaWjtUwz9sqsxtInMSyuErn36t8CD
Polarisation measurements
Instructions to install Madgraph, ROOT etc
The prerequisites for the MadGraph part of the tutorial are already installed on the same VM used for the Anomaly detection lecture.
If you don't want to use the VM you will need python 3, c++ and fortran compilers, plus the following programs and Python3 packages:
- Install MG https://launchpad.net/mg5amcnlo
- Install ROOT - Simplest way via conda https://anaconda.org/conda-forge/root/
- From MG prompt install Pythia8 and Delphes
>install pythia8
>install Delphes
- Install Jupyter https://jupyter.org/install
- Install some Python3 packages :
pandas, numpy, matplotlib, h5py
Also run a simple process in MG to make sure it runs through everything: you don't get errors and you get a ROOT file in PROC_sm_0/Events/. For example from the MG prompt run:
generate p p > e+ e-
launch
shower=Pythia8
detector=Delphes
Instructions to install LHAPDF and PHANTOM
1) Install and configure LHAPDF, following the instructions at this link:
https://lhapdf.hepforge.org/install.html (Quick-start instructions)
Note that you need to add LHAPDF in your path as an environmental path (section "Run-time search paths")
The latest version is 6.3.0.
To download a PDF set see the link https://lhapdf.hepforge.org/pdfsets.html, choose and wget the PDF set
wget http://lhapdfsets.web.cern.ch/lhapdfsets/current/NNPDF31_lo_as_0118.tar.gz
Place the tarball in the <path to LHAPDF>/share/LHAPDF/ folder and unzip it.
2) Extract the material from phantom tarball (tar -xvzf phantom_1_7.tar.gz) that you can find at this link:
https://www.dropbox.com/s/n88v3vdngtvk42g/phantom_1_7.tar.gz?dl=0
Enter the /phantom_1_7 directory.
You should see several fortran files (*.f, *.h), a makefile, an input card (r.in) and the directory /tools.
3) Edit the makefile script, according to the LHAPDF absolute path in your PC:
PDFLIBDIR = <path to LHAPDF>/lib
Note that the /lib folder is the one where you can find the libLHAPDF.so
4) run make (you may want to redirect the output of the compilation to a file, e.g. make > make.log &)
The compilation is rather long, e.g. ~40min, with a single CPU (Intel i5-3320M, 2.60GHz) and gcc 9.3.0 on Ubuntu 20.04.
If the compilation is successful the executable file phantom.exe is created.
If needed, note that make clean will clean all .o and .exe files.
5) To check that phantom_1_7 is correctly installed and interfaced with LHAPDF, edit the r.in file according to a PDF-set that is present on your PC (see above), e.g.
PDFname NNPDF23_nlo_as_0118_qed
6) Run ./phantom.exe
If the run starts smoothly, PHANTOM is properly installed and ready to use.
REFERENCES
(i) Ballestrero, Belhouari, Bevilacqua, Kashkan, Maina
"PHANTOM: a Monte Carlo event generator for six parton final states at high energy colliders"
https://arxiv.org/abs/0801.3359 [Comput. Phys. Commun. 180 (2009) 401-417]
original PHANTOM paper, polarization machinery NOT presented
(ii) Ballestrero, Maina, Pelliccioli
"W boson polarization in vector boson scattering at the LHC"
https://arxiv.org/abs/1710.09339 [JHEP 03 (2018) 170]
study of polarizations in W+W- scattering at the LHC with PHANTOM, polarization-selection methods presented
(iii) Ballestrero, Maina, Pelliccioli
"Polarized vector boson scattering in the fully leptonic WZ and ZZ channels at the LHC"
https://arxiv.org/abs/1907.04722 [JHEP 09 (2019) 087]
study of polarizations in W+Z and ZZ scattering at the LHC with PHANTOM
Programming in JULIA
The tutorial will follow the course at this repository:
https://github.com/oschulz/julia-course
Instructions on how to install the required packages can be found there, in the README file.
You should also install the EFTfitter.jl package. source and instructions are here:
https://github.com/tudo-physik-e4/EFTfitter.jl
