Particle Swarm Optimization OpenMPI

Table of Contents

Particle Swarm Optimization OpenMPI
- Build
  - Binary
  - Docker
  - Nix
  - Udocker
- Run
  - Local
  - Cluster (through PBS)
- Report
- Dependencies

ParticleSwarmOptimization-openMPI provides an basic implementation of PSO with support for cluster computation through OpenMPI, moreover it uses OpenMP for thread parallelization.

Long report	Short report	Presentation

Results

Sphere	Easom	Ackley

Build

Binary

make

Docker

make docker-build

with optional argument DOCKER_TAG

make DOCKER_TAG=mytag docker-build

Nix

nix build .

Udocker

make cluster-pull

Run

Local

./bin/particle-swarm-optimization pso-data.ini

or

mpirun -n <processesNumber ./bin/particle-swarm-optimization -u pso-data.ini

for optional arguments please execute ./bin/particle-swarm-optimization --help.

Cluster (through PBS)

Number of runs should be specified directly inside this script

make cluster-run

Report

make report

or

nix build .#report

Dependencies

If you have installed nix package manager

nix shell

otherwise dependencies are described below.

Build:

OpenMPI
sqlite C library
make
pkg-config
OpenMP

Python dependencies for jupyter analysis:

numpy
jupyterlab or jupyter notebook
pandas
matplotlib

Documentation and report generation:

pandoc
doxygen
latex with following packages:
- adjustbox
- babel-german
- background
- bidi
- collectbox
- csquotes
- everypage
- filehook
- footmisc
- footnotebackref
- framed
- fvextra
- letltxmacro
- ly1
- mdframed
- mweights
- needspace
- pagecolor
- sourcecodepro
- sourcesanspro
- titling
- ucharcat
- ulem
- unicode-math
- upquote
- xecjk
- xurl
- zref