Modeling Tools

We will rely on Solver in Excel as our modeling and solution tool.
Instructions on installing Solver.
Instructions for using Solver.
This is convenient in so far as everyone is familiar with Excel and so the fixed costs of installing, accessing and learning syntax for a new application are minimized. It is inconvenient in that:

  1. Solver is unreliable. It says models are not linear when they are. It says it has found a solution when it hasn't, etc.
  2. Solver combines the problem data with the model while algebraic modeling languages build a model abstraction and instantiate this with the data at the last moment
  3. Solver has rather tight restrictions on the number of variables.
  4. Solver is not an industrial strength tool.

So, I encourage you to learn an algebraic modeling language and implement the models we discuss in class in that language. Claudio is an excellent resource for help in developing these models.

Algebraic Modeling Languages: Some good resources include:

I prefer AMPL, but there are obvious advantages to free software and to having a full version available if your model exceeds the limits of the student version (XPress and AIMMS). Lindo, Lingo, What'sBest, Frontline's Solver etc. are acceptable, but barely so. I will provide AMPL models for many of the problems discussed in class.

 

Here are answers to AMPL questions: AMPL FAQS