Download
The latest version of the package can be obtained from GitHub (private).
# environment variable
export GIT_EMAIL="your email address"
export GIT_USER="your account name"
export GIT_TOKEN="your token"
# download package
git clone https://${GIT_USER}:${GIT_TOKEN}@github.com/#####
Note
DASH is not an open source. Please contact Dr. Yoshiaki Abe for possible collaborative research.
Requirements
Python3
python3.6 or later is used in various process.
LAPACK
LAPACK is used to process matrix operation in fortran program (only in RBF).
git clone https://github.com/Reference-LAPACK/lapack.git
cd lapack
cp make.inc.ref make.inc
make -j 10
change Initial/03CreateFSIMat/src/RBF/Makefile to add lapack directory
Doxygen
Doxygen is used to generate documents for C language codes. If you do not need documents, Doxygen can be skipped.
Build
Exeutable programs are built automatically by “setup.sh” Before executing “setup.sh”, check your computing system.
local (workstation)
If using local machine, no change is required.Execute “setup.sh”.
bash setup.sh
AFI-NITY (HPC in IFS)
For AFI-NITY, please modify “envs.sh” as written below.
ComputeSystem=AFI
On AFI, intelpython (Python3.7.7) does not work with OpenMPI/gcc. But, “module load intel” has set the python3 to intelpython. So, please install Python3.6.8 or Python3.7.5 (both were confirmed) and choose the following example (alias or overwrite PATH to python3)
# Example 1
alias python3=$HOME/local/Python-3.6.8/bin/python3
# Example 2
export PATH=$HOME/local/Python-3.7.5/build/bin/:${PATH}
Now you can execute “setup.sh”.
bash setup.sh
Fugaku (HPC of Riken)
This contents will be updated…
If using Fugaku, please modify “envs.sh” as written below.
ComputeSystem=fugaku
In Fugaku, the python environment can only be built on compute nodes. At first, Comment out commands except python and execute “setup.sh” in interactive job.
pjsub --interact -S -L "node=1" -L "elapse=00:15:00" --sparam "wait-time=3600" -g #### -x PJM_LLIO_GFDCACHE=#### -x PJM_LLIO_GFSCACHE=#### -N "d_env"
bash setup.sh
Next, comment out the description of python in “env.sh” and execute “setup.sh” on the login node.
bash setup.sh