optimeed 2.2.1

Optimeed aims at easing data generation and visualization. Its key strength is its adaptability to any problem thanks to its flexibility, while accounting for parallelization, disk space and RAM usages.

It is divided in 4 Modules:

• core Embeds all the tools to load/save python objects in plain text using json-like format.
• optimize A framework to solve optimization problem, while profiting from core saving utilities.
• consolidate A framework to perform sensitivity analyses, while profiting from core saving utilities.
• visualize A framework for data analysis. Provides GUI for optimize and consolidate, using 2D interactive graphs. For the optimization, results can be displayed in real time.

Requirements

• PyQt5 for visualisation -> pip install PyQt5
• pyopengl for visualisation -> pip install PyOpenGL
• Numpy -> pip install numpy
• Optional
  • pandas which is only used to export excel files -> pip install pandas
  • nlopt library for using other types of algorithm. -> pip install nlopt
  • inkscape software for exporting graphs in .png and .pdf)
  • plotly library for 3D plots. -> pip install plotly

Installation

To install the latest optimeed release, run the following command:

pip install optimeed

To install the latest development version of optimeed, run the following commands:

git clone https://git.immc.ucl.ac.be/chdegreef/optimeed.git
cd optimeed
python setup.py install

Support

Documentation optimeed

or

Gitlab (preferably), read the guided tutorials.

or

Send mail at christophe.degreef@uclouvain.be.

License

The project is distributed "has it is" under GNU General Public License v3.0 (GPL), which is a strong copyleft license. This means that the code is open-source and you are free to do anything you want with it, as long as you apply the same license to distribute your code. This constraining license is imposed by the use of Platypus Library as "optimization algorithm library", which is under GPL license.

It is perfectly possible to use other optimization library (which would use the same algorithms but with a different implementation) and to interface it to this project, so that the use of platypus is no longer needed. This work has already been done for NLopt, which is under MIT license (not constraining at all). In that case, after removing all the platypus sources (optiAlgorithms/multiObjective_GA and optiAlgorithsm/platypus/*), the license of the present work becomes less restrictive: GNU Lesser General Public License (LGPL). As for the GPL, this license makes the project open-source and free to be modified, but (nearly) no limitation is made to distribute your code.