humam/6/28/2015
- made a couple of changed to the adaptive order algorithm for the parallel implementation to fix the slow performance.
- thought about better ways to distribute the work load among worker processes, and found 2 ways to optimally distribute the load. I will change the code to use these better ways later.