This project was undertaken as part of CSE 6140 - CSE Algorithms. The final report for the project can be found here.

• Parul Awasthy ([email protected])
• Young Jin Kim ([email protected])
• Ankit Srivastava ([email protected])

SCons is used for building the project. It is based on Python and can be easily downloaded and installed. After installing SCons, following needs to be executed on the command line:

scons [DEBUG=1]

This builds the project in builds/[release|debug] directory under the current directory and installs the final executable, named tspsolver[_debug], in the current directory.

Alternatively, the project can also be built using Makefile by executing following on the command line:

make

This builds the project in the top level directory under the current directory and installs the final executable, named tspsolver, in the current directory. Note that only release version can be built using Makefile. Please use SCons, described above, if you want to build the debug version.

Boost libraries are required for building the project. Installing libboost-all-dev, on Linux, should do the trick.

The executable can be run as follows:

./tspsolver[_debug] --alg [BnB | Approx | Heur| LS1 | LS2] --inst <filename.tsp> [--time <cutoff (in seconds)>] [--seed <seed number>]