Software

On this page you will find scripts/codes/software I have written.

COLA (Conic Optimization with Linear Approximations) is a solver written in C++ that solves second order conic optimization (SOCO) problems. It uses outer approximation to solve SOCO problems. It depends on COIN-OR’s solver interface OSI and lp solver CLP.

DietCOLA (Discrete COLA) is a solver for discrete SOCO problems. For now it uses a naive branch and bound method. It depends on COIN-OR’s ingenious search library ALPS (Abstract Library for Parallel Search). By default it uses COLA to solve continuous relaxation problems, Mosek can also be used for this through interface provided. Nowadays I spent most of my time to improve DietCOLA, more improvements will be coming, stay tuned.

Check GiMPy (Graph Methods in Python) and GrUMPy (Graphics for Understanding Mathematical Programming) packages for Python. GiMPy contains a graph class and associated methods for visualizing various graph-based algorithms. GrUMPy is build on top of GiMPy and used for visualizing branch and bound tree of integer linear programs. The purpose is to build and visualize the tree on the fly when the solver is working, for observing/analyzing the process. These packages are available through Python Package Index (PyPI) and Coin-OR website. You can install them using easy install.

Check BLiMPy (Basic List Implementation in Python) package. It is a linked list implementation. We developed this to use in GiMPy.

Turing machine simulator is a C++ class that simulates a Turing Machine. Turing machine class has a very simple API that provides writing Turing machine programmes. It is based on the Turing machine simulator written in Python. The python version can be found in http://code.activestate.com/recipes/252486-turing-machine-simulator.

Unlimited Register Machine (URM) is a model of computation, like Turing Machine. Computability theorists like to work with URMs, at least Nigel Cutland does. I took a reading class based on Cutland’s “Computability: An Introduction to Recursive Function Theory” book. I wrote a Python class that simulates URMs, for fun. I implemented some of the example programs/exercises from the book. You can get URM simulator and examples from here. Have fun with URM programming!

Salbp is a python class that solves Simple Assembly Line Balancing Problems (SALBP) of type 2. It depends on some other open-source softwares. For further information about the problem or how to use Salbp you can check my qualifier report which is about SALBP type 2. Slbp-1.0 misses solution retrivial methods. You can use Pulp methods for this.