Sylvaticus,
I find your approach very interesting, particularly the flexibility to get around CPU or memory. I am still establishing which will be my limitation. At the moment my problem takes 7 minutes to run through the 9,000 iterations, writing the *.dat file as it goes. I have not tried loading all of the outputs into a matrices, and then printing out after the main loop has completed. I don’t think there will be a memory problem. Ultimately I want the warp the whole program up in an optimization engine, which will be CPU limited. During these runs I would suppress writing of most of the output results until the optimum solution is found.
Thanks for you ideas
Peter