Setting mamba environments
General jupyter environment
mamba create --name 'myenv'
mamba activate myenv
mamba install jupyter
mamba install matplotlib numpy scipy meshio
mamba install -c conda-forge ipympl
mamba install jupyter_contrib_nbextensions
jupyter nbextension enable ipympl --py --sys-prefix
jupyter nbextension install --py --sys-prefix --overwrite ipympl
SLaSi, CPU version
Prerequisites:
mamba create --name "slasi-cpu" python=3.9 -y
mamba activate slasi-cpu
mamba install gcc gxx openmpi cmake cunit zlib backports.lzma sphinx -y
Preparing SLaSi libraries:
cd slasi-cpu/libsources
make python nlopt -j4
# optionally, one may remove unnecessary source folders here
cd ..
make -f Makefile.docs -j4
cd ..
mkdir slasi-build-cpu
cd slasi-build-cpu
cmake ../slasi -DPARALLEL=MPI && make -j4
SLaSi, GPU version
Prerequisites:
mamba create --name "slasi-nv" python=3.9 -y
mamba activate slasi-nv
mamba install gcc gxx openmpi cmake cunit zlib backports.lzma sphinx -y
mamba install cuda-toolkit cuda-version=12
mamba install cuda-cudart cuda-version=12
#mamba install cuda-toolkit cuda-version=11 -y
#mamba install cuda-cudart cuda-version=11 -y
mkdir slasi-build-nv
cd slasi-build-nv
cmake ../slasi -DPARALLEL=CUDA && make -j4
ln -s ./slasi ./slasi-nv