TSP Software

Abstract

In the preceding chapters we have seen various algorithms for solving the Traveling Salesman Problem (TSP) and other related problems either by computing a provably optimal solution or by computing an approximate (heuristic) solution. Results of extensive comparative studies of various competitive heuristic algorithms for the symmetric and asymmetric traveling salesman problems are presented in Chapters 9 and 10. Most often these comparative studies focus on two important criteria — the quality of the solution produced and the computational time. While computational performance is one of the most important criteria, several other desirable criteria should to be considered in evaluating a software for any computational problem. An academician may be interested in a software with the aim to illustrate various features of an algorithm and a practitioner may want to solve quickly a small problem without bothering to work with a large and sophisticated source code. Moreover, due to the increasing use of Internet and World Wide Web in the academia, Java applets illustrating various steps of an algorithm have pedagogical advantages. Even in the case of source codes, one may prefer a specific programming language to another, whether it is a general-purpose language such as FORTRAN, Pascal, C, C++, etc. or it is a special-purpose language such as AMPL, MAPLE, Mathematica, etc.