Code that I've written

Codes to find minimum cuts

The codes can be found here and input files can be found here . You can find a link to both a short (10 page) and long (130 page) paper about these codes here These sources correspond to what is described in Matt Levine's masters thesis. We reorganized and cleaned up quite a bit. K is now much better, with more heuristics and dynamic trees, using a dynamic tree implementation by Tamas Badics. We also made sure that HO, NI, and K will return the cut itself, as well as the value.

code for concurrent multicommodity flow

This is the code of Leong, Shor and Stein. It is described here
Here is code for undirected graphs.
Here is code for directed graphs.

Hard instances for RELAXT-III:

These files (in DIMACS format, tarred, compressed) are instances of a minimum-cost flow problem on a particular 49 node graph. The instances differ in the choice of capacities and demands. We ran these problems using the RELAXT-III code of Bertsekas and Tseng. The interesting thing that happened was the running times varied greatly. A file with the name gte_bad.XXXX represents a graph which took XXXX milliseconds ( on a circa 1990 computer). These seems to be bad examples for the RELAXT-III code. Any good explanations? These problems came about as subproblems in the multicommodity flow code of Leong, Shor, and Stein.