December 22, 2002

Another Error Calculation Error

I found another big problem in the automatic policy utility error calculations. Instead of returning an error value between -1.0 and +1.0 (e.g. +/- 0.05 for 5% error), I was returning error values one hundred times greater. I was coding the error margins incorrectly as integers instead of the fractions they should be. So my +/- 50% error was adding or subtracting up to 50 from the utility value which is only supposed to range from 0.0 to 1.0 (or -0.5 to +1.5 with a 50% error range). So once again, I'm rerunning the simulation.

Posted by josuah at December 22, 2002 12:05 AM UTC+00:00