Visualizing Branch-and-Bound Algorithms

We present a suite of tools for visualizing the status and progress of branch-and-bound algorithms for mixed integer programming. By integrating these tools with the open-source codes CBC, SYMPHONY, and GLPK, we demonstrate the potential usefulness of visual representations in helping a user predict future progress of the algorithm or analyzing the algorithm’s performance. We have also implemented a flexible toolkit, called the Branch and Cut Analysis Kit, that can be used in conjunction with any instrumented solver to create visual representations in the form of image files. The kit will be made available under an open-source license.