==== Install Madgraph ====
<alert type="info">**Note:** There can be troubles in compiling/library linking/etc at the runtime.

</alert>
Initial setup - install MG5 from launchpad
<code>
mkdir -p ~/work/MG5
cd ~/work/MG5
wget https://launchpad.net/mg5amcnlo/2.0/2.6.x/+download/MG5_aMC_v2.6.4.tar.gz
tar xzf MG5_aMC_v2.6.4.tar.gz
</code>

Setup the devtoolset to use updated gcc and python.
<code>
# Don't need to source following lines in Ubuntu and Mac.
# source /opt/rh/devtoolset-4/enable
# source /opt/rh/python27/enable
cd MG5_aMC_v2_6_0
</code>

Install missing tools. Order does matter.
<code>
bin/mg5_aMC
#MG5_aMC> install update
MG5_aMC> install zlib; install boost;
MG5_aMC> install lhapdf6; install hepmc;
MG5_aMC> install oneloop; install ninja;
MG5_aMC> install pythia8
MG5_aMC> install collier
MG5_aMC> quit
</code>

<alert type="info">For NLO production, the "OneLoop" package have to be recompiled, after modifying the source code</alert>
Open the fortran source code ''HEPTools/oneloop/OneLOop-3.6/src/avh_olo_print.f90'' and modify the line 61, replace ''real(kindr2)'' to ''double precision'' (in the ''printr'' function).
This is because of a gfortran bug. See https://bugs.launchpad.net/mg5amcnlo/+bug/1631604 and https://answers.launchpad.net/mg5amcnlo/+question/268548#16
<code>
  double precision :: xx     !|RCTYPE=intrinsic
!  real(kindr2) :: xx     !|RCTYPE=intrinsic
</code>

and compile the package

<code>
cd HEPTools/oneloop/OneLOop-3.6
./clean.sh
./create.py
\cp libavh_olo.a ../
</code>

<alert type="info">To use the pythia8, you need additional step to link missing libraries</alert>
First step is to build missing library ''boost-iostreams''.

<code>
if [ ! -f HEPTools/lib/libboost_iostreams.so ]; then
  cd HEPTools/boost/boost_1_59_0
  ./bootstrap.sh --with-libraries=iostreams
  ./b2
  \cp stage/lib/* ../lib/
  cd ../..
  cd lib
  for i in ../boost/lib/libboost_iostreams*; do ln -s $i; done
  cd ../..
fi
</code>

Then add the following lines to fix build flags https://answers.launchpad.net/mg5amcnlo/+question/659847
<code>
sed -i "s;PYTHIA8_PATH=NotInstalled;PYTHIA8_PATH=`pwd`/HEPTools/pythia8;g" Template/LO/Source/make_opts
sed -i 's;OUT+=" -ld";OUT+=" --libs";g' HEPTools/bin/pythia8-config
sed -i 's;OUT+=" -ld";OUT+=" --libs";g' HEPTools/pythia8/bin/pythia8-config
</code>

enable lhapdf6 for all of your generation (download other PDFSets depending on your needs)
<code>
cd HEPTools/lhapdf6
bin/lhapdf get NNPDF23_lo_as_0130_qed
</code>

Then you will be able to start the generation.
<code>
./bin/mg5_aMC
MG5_aMC> generate p p > t t~ [QCD]
MG5_aMC> output ttNLO
MG5_aMC> launch
</code>

<alert type="info">To run pythia8 in launch level, you edit runcard; replace HERWIG6 to PYTHIA8.</alert>

You can rerun the shower program.
<code>
cd ttNLO
bin/shower run_01
</code>

If you run the MG5 at the leading order, LHAPDF is required during the runtime, related on systematics. Set the environmental variables before running the mg5_aMC (or put into your .bashrc file) to prevent such a problem.
<code>
export PYTHONPATH=YOUR_MADGRAPH_PATH/HEPTools/lhapdf6/lib64/python2.7/site-packages:$PYTHONPATH
export LD_LIBRARY_PATH=YOUR_MADGRAPH_PATH/HEPTools/lhapdf6/lib:$LD_LIBRARY_PATH
</code>

To speed up the parton shower step, you can set ''nsplit_jobs=10'' in the shower_card.dat

==== Running on Cluster mode ====
The default setup is ''multicore-mode'', but also you can use the''cluster mode'' if you have to produce lots of events.
Edit the ''madgraph/various/cluster.py'' file first (line 1286), since the default submission script is not fit with the HTOP cluster. (see https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/General-Scripts#no1)

<code>
        command = ['qsub','-o', stdout,
                     '-S', '/bin/bash',
                     '-N', me_dir,
                     '-e', stderr,
                     '-V']
</code>

and add the following lines to your configuration file, ''input/mg5_configuration.txt''
<code>
cluster_type = sge
cluster_size = 96
</code>

and run with ''--cluster'' option
<code>
bin/generate_events --cluster
</code>
But in this case, you should not run delphes automatically because multiple delphes sessions will make a crash.

==== Basic Syntax ====

==== Get New PDF ====

==== Parameters for Run Card ====

==== Disable HEPMC compression ====
만들어진 HEPMC파일을 압축하는데에는 시간이 오래 걸리므로, 외부에서 delphes를 사용할 것이라면 압축하지 않고 다음 job으로 넘어가는 것이 빠르다.
<code>
vi path_to-MG5/madgraph/interface/madevent_interface.py
</code>
line 5291의 ''misc.gzip(file_path,stdout=file_path)''를 추석처리한다. 이미 데이터셋 'output folder_name'를 만들었다면, 해당 폴더에서
<code>
vi bin/internal/madevent_interface.py #코드 수정
rm bin/internal/madevent_interface.pyo
</code>