This is a program implemeting the Particle Swarm Optimisation (PSO) algorithm for searching the global minimum of a series of clusters placed on top of surfaces.
To quickly install KCL_PSO, we encourage to do it inside a virtual environment, which can be achieved in the following way:
- Create a directory named as you want and access it (in this case called work_pol):
[~] mkdir work_pso
[~] cd work_pso
- IN CASE YOU ARE INSTALLING THE PACKAGE IN YOUNG You need first to load these modules BEFORE INSTALLING THE LIBRARRY. To do so, please type the following commands in your session:
module purge
module load gcc-libs/4.9.2
module load openblas/0.3.7-serial/gnu-4.9.2
module load python3/3.8
- Create a virtual environment named pol (this name can be changed, of course) and activate the environment:
[~] python -m venv kcl_pso
[~] source kcl_pso/bin/activate
- Get KCL_PSO from the GitHub repository:
git clone https://github.com/alejandrosantanabonilla/PSO.git
- Install KCL_PSO in this folder using the virtual environment
[~] cd PSO
[~] pip install .
To test the code, there is a folder inside the PSO directory called test. Inside there is a folder called input_files, where all the corresponding VASP files EXCEPT POTCAR are located. Please add POTCAR in that directory before starting any calculation.
You can submit a first test, just by typing:
qsub job.sh
This should work. To check your job, you can type
qstat -u mmmXXXX
and look for pso_calculation job.