Pyomo – Optimization Modeling in Python

by William E. Hart, Carl D. Laird, Jean-Paul Watson

This book provides a complete and comprehensive reference/guide to Pyomo (Python Optimization Modeling Objects) for both beginning and advanced modelers, including students at the undergraduate and graduate levels, academic researchers, and practitioners. The text illustrates the breadth of the modeling and analysis capabilities that are supported by the software and support of complex real-world applications. Pyomo is an open source software package for formulating and solving large-scale optimization and operations research problems. The text begins with a tutorial on simple linear and integer programming models. A detailed reference of Pyomo's modeling components is illustrated with extensive examples, including a discussion of how to load data from data sources like spreadsheets and databases. Chapters describing advanced modeling capabilities for nonlinear and stochastic optimization are also included. The Pyomo software provides familiar modeling features within Python, a powerful dynamic programming language that has a very clear, readable syntax and intuitive object orientation. Pyomo includes Python classes for defining sparse sets, parameters, and variables, which can be used to formulate algebraic expressions that define objectives and constraints. Moreover, Pyomo can be used from a command-line interface and within Python's interactive command environment, which makes it easy to create Pyomo models, apply a variety of optimizers, and examine solutions. The software supports a different modeling approach than commercial AML (Algebraic Modeling Languages) tools, and is designed for flexibility, extensibility, portability, and maintainability but also maintains the central ideas in modern AMLs.

238 pages, Hardcover

First published January 1, 2012

Reviews

[calavera · Rating 3 out of 5]

Pretty strange organization of the book with related things, that should be mentioned together, put into separate chapters. There are things omitted from both the Pyomo documentation and this book.

However, Pyomo and PySP are very useful and FREE tools that can, in some part of most usual operations, replace pricey AMPL and GAMS. Of course Pyomo supports just a subset of most popular modeling languages but is easily extendible due to its embeddedness into the general purpose language and learning curve is less steep for people familiar with python.

This book is, despite its noticeable shortcomings, useful in the first steps with Pyomo and later as a reference manual.