dissim - A discrete simulation optimization package

dissim is a Python package for discrete simulation optimization. It provides algorithms for solving optimization problems where the variables are discrete and the objective function is estimated through simulation. This can be useful in situations where the objective function is difficult to evaluate analytically or where the system being optimized has stochastic elements.

Algorithms

The package currently includes the following algorithms:

Adaptive Hyperbox Algorithm

This algorithm is an instance of a locally convergent, random search algorithm for solving discrete optimization via simulation problems. It is more efficient in high-dimensional problems compared to the COMPASS algorithm. This algorithm is described in the research paper “An adaptive hyperbox algorithm for high-dimensional discrete optimization via simulation problems” by Xu, Jie, Barry L. Nelson, and L. Jeff Hong.

Stochastic Ruler Algorithm

This algorithm was originally described by Yan and Mukai in 1992 for asymptotically determining the global optima of discrete simulation optimization problems. It has been proven to have asymptotic convergence in probability to the global optimum solution. This algorithm is described in the research paper “Stochastic discrete optimization” by Yan, Di, and H. Mukai.

Simulated Annealing Algorithm

The Simulated Annealing Algorithm, introduced in 1983 by Kirkpatrick, Gelatt, and Vecchi, revolutionized the realm of optimization for tackling complex, discrete simulation optimization problems. This elegant approach has demonstrated remarkable asymptotic convergence properties, reliably guiding its search towards global optima. This algorithm is described in the paper "Optimization by Simulated Annealing" by S. Kirkpatrick, C. D. Gelatt Jr., and M. P. Vecchi.

Requirements

To use Dissim, you will need to have the following packages installed in your system: numpy scikit-learn pandas dask matplotlib You can install these packages using pip by running pip install numpy scikit-learn pandas dask matplotlib

Installation

To install dissim, download the dissim-0.1.3-py3-none-any.whl file from the root folder of the repository and run pip install ./dissim-0.1.3-py3-none-any.whl in the same folder.

Usage

Example files for using the algorithms in this package can be found here and here.

Tree

/c/Users/hp/dissim
|-- LICENSE.txt
|-- README.md
|-- build/
   |-- bdist.win-amd64/
      |-- *
   |-- lib/
      |-- dissim/
         |-- __init__.py
|-- codes/
   |-- algorithms/
      |-- Adaptive_Hyperbox_Algorithm/
         |-- AdaptiveHyperbox.py
         |-- examples/
            |-- e1.py
            |-- e2.py
            |-- e3.py
            |-- e4.py
            |-- e5.py
            |-- e6.m
            |-- e6.py
      |-- Simulated_Annealing/
         |-- SA_CRN.py
         |-- SimulatedAnnealing.py
         |-- examples/
            |-- e1.py
            |-- e2.py
            |-- e3.py
            |-- e4.m
            |-- e4.py
            |-- e5.py
            |-- multinodal_plot.py
         |-- func.m
         |-- use_mrg.py
      |-- Stochastic_Ruler_Algorithm/
         |-- SR_N1N2.py
         |-- StochasticRuler.py
         |-- examples/
            |-- e1.py
            |-- e2.py
            |-- e3.py
            |-- e4.py
            |-- e5.py
   |-- test cases/
      |-- hotel.py
|-- dissim/
   |-- __init__.py
|-- dissim.egg-info/
   |-- PKG-INFO
   |-- SOURCES.txt
   |-- dependency_links.txt
   |-- requires.txt
   |-- top_level.txt
|-- dist/
   |-- dissim-0.1.2-py3-none-any.whl
   |-- dissim-0.1.2.tar.gz
   |-- dissim-0.1.3-py3-none-any.whl
   |-- dissim-0.1.3.tar.gz
|-- setup.py

Contact Us