Read my latest posts!

Wednesday, May 11, 2005

A word about the Classification Script

After making the change to the script correcting the sense of the vigilence test, I made a test sheet of some 31000 rows of data with the maximum of 31 columns initialized with random boolean values. I am currently running the script on that test data, and it isn't pretty thus far. It has been running with almost complete utilization of the CPU for well over 67 hours, and still has not completed. If you choose to use this script, I would suggest reduction of the number of iterations. I imagine there is a lot of resonance going on, and if it never settles down, then the script must run every iteration. I coded for the occurence that if there is no changes and the iterations have not completed, termination of the script is called early, but alas, this does not appear to be the case for the test data. I will update how long it takes.after it completes.

Update: Alright already. After running the G*@3#~! script for over 177 hours, I killed it. Geesh. Granted it was only using 50 Megs of memory, but the processor usage was getting to a pain in the >insert choice of anatomy here<. Looks like something for me to convert to C#.

Labels: , ,


Post a Comment

<< Home