Tuesday, June 23, 2015

disambiguation

upon really fine examination of all the parts of the Ga program I found that two major parts of the program where not quite correct. The effects where seen when the numbers that the computer dealt with and the numbers that where presented in the log files where not matching up. It turns out that when the generation was being read to make the next generation the float arrays that where being altered where for the next generation was them selves being altered. ... ugg references...(-U_0-)  Meaning that if a individual had a good fitness, it would share its genetic code, then have that code be altered to a better or worse fitness and the program would not notice. This could have lead to all the default values. I'm going to run the program again and see if It fails again.

No comments:

Post a Comment