NPGrowth_V1 — Installation Guide

Requirements

• Conda (Miniconda or Anaconda)
• Git
• CMake ≥ 3.16
• A C++ compiler with MPI support (e.g. OpenMPI via Homebrew on macOS)
• Python 3.12

---

Step 1 — Clone the NPGrowth_V1 repository

git clone https://github.com/nanoMLMS/NPGrowth_V1.git
cd NPGrowth_V1

---

Step 2 — Create and activate the Conda environment

conda create -n NPGrowth_V1 python=3.12
conda activate NPGrowth_V1

---

Step 3 — Clone LAMMPS

From inside the NPGrowth_V1 directory:

git clone https://github.com/lammps/lammps.git
cd lammps
git checkout stable

---

Step 4 — Replace the modified LAMMPS source files

NPGrowth_V1 requires a patched version of fix_deposit that supports the target keyword. Copy the modified files into the LAMMPS source tree:

cp ../lammps_deposition_modified/fix_deposit.cpp src/
cp ../lammps_deposition_modified/fix_deposit.h src/

Fix API compatibility

The patched file uses an older minimum_image API. Update it to match the current LAMMPS version:

sed -i '' 's/domain->minimum_image(delx,dely,delz)/domain->minimum_image(FLERR, delx, dely, delz)/g' src/fix_deposit.cpp

Note (Linux): omit the '' after -i:

sed -i 's/domain->minimum_image(delx,dely,delz)/domain->minimum_image(FLERR, delx, dely, delz)/g' src/fix_deposit.cpp

Verify the fix was applied (should show 2 lines with FLERR):

grep "minimum_image" src/fix_deposit.cpp

---

Step 5 — Build LAMMPS

mkdir build
cd build

(on indaco:
module load intel-oneapi-mpi
module load intel-oneapi-mkl

cmake ../cmake \
  -DCMAKE_INSTALL_PREFIX=$HOME/lammps \
  -DBUILD_MPI=on \
  -DBUILD_OMP=off \
  -DPKG_MOLECULE=on \
  -DPKG_MANYBODY=on \
  -DPKG_KSPACE=on \
  -DFFT=MKL \
  -DCMAKE_C_COMPILER=mpicc \
  -DCMAKE_CXX_COMPILER=mpicxx


)

cmake ../cmake \
  -DCMAKE_INSTALL_PREFIX=$HOME/lammps \
  -DPKG_MOLECULE=on \
  -DPKG_MANYBODY=on \
  -DPKG_KSPACE=on \
  -DBUILD_MPI=yes

make -j 8
make install

The compiled binary will be installed at $HOME/lammps/bin/lmp.

---

Step 6 — Run a sample simulation

Navigate to the sample directory and launch LAMMPS with MPI:

cd ../../sample
mpirun -np 4 $HOME/lammps/bin/lmp -in sample.in

Adjust -np 4 to match the number of CPU cores available on your machine.

---

Directory structure

NPGrowth_V1/
├── code/                          # Python scripts
├── lammps/                        # LAMMPS source (cloned in Step 3)
├── lammps_deposition_modified/    # Patched fix_deposit files
│   ├── fix_deposit.cpp
│   └── fix_deposit.h
├── sample/                        # Example simulation
│   ├── sample.in                  # Main LAMMPS input script
│   ├── molecule0.txt
│   ├── potentials/
│   ├── make_deposit/
│   └── seeds/



## To run:

To create the configuartion for the deposition:

python3 make_deposition.py \                                                                                                   
  --config config/single_gun_single_target.json \
  --output output/test_0

then to run the lammps:

mpirun -np 4 $HOME/lammps/bin/lmp -in sample.in