Internet becomes fueling ground for Open Source mathematics discussion

An interesting post from Timothy Gowers, Field's medalist and Fellow at Trinity College, suggests starting an online weblog for publically collaborating mathematical concepts and problem solving. Glyn Moody, on his blog Open..., took it one step further and suggested that this could be considered Open Source Mathematics. Is it possible to use the Open Source movement as a methodology of solving mathematical principles and problems? I absolutely believe this is possible. In fact I am willing to suggest that this has already been in existence.

Math is a language that has no race, religion, or region. Math has been collaborated by all sorts of people, in academia or not, for centuries. The only thing that has kept public collaboration is the media by which it is presented to the masses. In the past the methodology was journals, papers, lectures, and small forums of people to collaborate. Now the age of the internet has brought a new media to present mathematical ideas. Journals are already being spread online. Blogs and online chat rooms have brought like-minded interest groups together. I believe we could be in a new age of mathematical revolution.

Why go proprietary when you can go Free part 2

Since I last posted about Free software tools I have found other great Industrial Engineering tools. Statistics plays a major role for the Industrial Engineer and Operations Research analyst. There are a huge amount of statistical tools in the market for computers and software. I have found that there is numerous Free Software tools available for analyzing statistics and data sets.

Here is a Free Software equivalents table for statistical software

S, SAS --> R-Project

Matlab, Scilab --> SciPy (requires Python programming language)

SPSS --> PSPP

Why go proprietary when you can go Free

The tools in the Industrial Engineer and Operations Research tool bag is increasing with the advances of software systems. In the past the tools were relegated to proprietary software that would cost an organization licensing fees, software updates, and support services. Yet there is a shift in thinking in how software is developed, delivered, and administered to the masses. That shift is in free and open source software. This free software movement has shifted the thinking that good software need not be copyrighted and licensed for proprietary distribution. Instead software should be made freely to the masses to use, distributed, and even improve upon.

The free and open source software movement has even made its way to Industrial Engineer and Operations Research tools. There are many proprietary equivalents that can be found in the free software realm. Most of which will even work on a Windows platform. I have listed some of these software tools in a equivalents table.

Spreadsheet
Microsoft Excel ---> OpenOffice Calc, KOffice KSpread

Spreadsheet LP Solver
Excel Solver ---> OpenOffice.org Calc Optimization Solver

Math Programming Software
AMPL, CPLEX, GAMS ---> GLPK, OpenOpt


These are but only a few equivalents. You can see more free and open source equivalents to all sorts of software on the following links.

The Table of Equivalents
LinuxQuestions.org equivalents table

What is Operations Research?

When I tell people I have a masters degree in Operations Research they often say "What is Operations Research?". Michael Trick has a great blog post examining this very question of explaining Operations Research to others. In fact he even solves that question with the ever so appropriate Operations Research tool of using an algorithm.

My typical response to the question "What is Operations Research" is using mathematics and statistics to help solve a business decision. Better yet, from the blog article, is Jim Orlin's response. Jim states Operations Research is "science of better decision making". Very good response and is one I will use in the future.

OpenOpt releases version 0.21

The developers of OpenOpt have released the newest version 0.21 to the masses on December 15. Main developer, Demitry, has listed some improvements such as NLP and NSP algorithm changes, converts between optimization problems, and an new OpenOpt GUI.

OpenOpt is a free and open source optimization tool kit software package. It is specifically developed for Python software language. Several models can be built with the OpenOpt kit including Linear, Quadratic, Non-linear, and Mix Integer Progamming problems. The purpose of OpenOpt is to bridge the gap between strong, powerful optimization software and rapid development and implementation.

OpenOpt has a new forum and location for their latest updates. Please follow this link to the Numerical Optimization Forum at OpenOpt.org.

The alternative Calc Optimization Solver

For those industrial engineers that use optimization tools for operations research you no doubt know about the Microsoft Excel Solver. An optimization tool embedded into a spreadsheet can be a real life saver. Linear programming problems can be developed rather rapidly. Concepts can be quickly tested. Applications can be delivered on a software platform that is rather ubiquitous. Yet is there an open source free software solution available?

Of course there is one available. The Calc Optimization Solver is available for use with the Open Office software suite. The modeler will no doubt recognize the similarities to the MS Excel Solver. Calc's Optimization Solver can tackle most linear programming scenarios. Problems can be developed by assigning Calc cells to variables or equations. Integrality can be set if an integer results are desired. Non-negativity can be set by just setting a flag in the Options menu.

Calc's Optimization Solver may not be the desired tool for serious optimization problems yet it can be real handy for most situations. Serious applications for business use can be developed with the Optimization Solver tool. The Calc Optimization Solver can also be used to help teach operations research lessons to peers and pupils.

More free software tools for Operations Research

I came across a site while looking for more GLPK (Gnu Linear Programming Kit) information. The site is Decision Tree for Optimization Software.

Decision Tree for Optimization Software

The site is great information for different types of software related to different types of problem sets such as Linear Programming, Non-linear programming, Multiobjective, Least Squares, etc. One could spend hours on this site alone just trying to draw conclusions for the type of software engine to solve an optimization problem.