sciss / linkernighantsp

An implementation of the Lin-Kernighan heuristics algorithm for the (symmetric) traveling salesman problem (TSP).

Version Matrix


Build Status Maven Central


An implementation of the Lin-Kernighan heuristics algorithm for the (symmetric) traveling salesman problem (TSP).

This is derived from – using sbt for building, and translating to the Scala programming language. Performance has been monitored to be somewhat similar to the original code (for a graph with 734 nodes, we measured a 13% slow-down compared to the Java sources). The project has been modified to use a generic cost table instead of 2-dimensional cartesian coordinates, and the initial tour has to be explicitly provided.

building and running

This project builds with sbt against Scala 2.13, 2.12, Dotty (JVM) and Scala 2.13 (JS).

To compile, use sbt compile, to run the demo use sbt run. You can also provide a dataset number, e.g. sbt 'run 17'.


To use this project as a library, use the following artifact:

libraryDependencies += "de.sciss" %% "linkernighantsp" % v

The current version v is "0.1.3"

Below is the original read-me.


It implements exactly the same features described by Shen Lin and Brian Kernighan in their original paper "An Effective Heuristic Algorithm for the Traveling-Salesman Problem"

In addition, this implementation reads .tsp files to form a representation of the TSP problem; however, the format of the .tsp file is not yet TSPLIB compliant, but it would be a nice and easy issue to fix.

How to contribute

We want to keep contributions as simple as possible, so the rules are simple: you contribute changes, I review them, if I agree with the changes I will pull it down.

That being said, there are a few recommendations for the quality of code to contribute:

  • Efficiency: make the code as efficient as possible, efficiency is one of the key aspects of this kind of algorithms.
  • Readability: document all the new functions by adding comments, and if there are tricky pieces of code, comment them as well.
  • Reliability: ensure the code works!

Finally, review the TODO list in this page.

How to run

Currently, the code is ran by compiling all the files and running the Main class.


  • Original Paper by S. Lin and B. W. Kernighan

Lin, Shen; Kernighan, B. W. (1973). "An Effective Heuristic Algorithm for the Traveling-Salesman Problem". Operations Research. 21 (2): 498–516. doi:10.1287/opre.21.2.498.

  • Waterloo TSP test data


NOTE: we are currently using a different format for our test data.