I mean that if an individual returns NaN, then set it to a very poor fitness. Try printing out a note whenever this happens, and you'll see that very quickly NaN-generating individuals disappear from the population. Sean On Jul 25, 2005, at 2:40 PM, Steve Butcher (Steve-O) wrote: > Sean, > > Thanks for the reply...I got the gist of most of it. However, your > parting comment was a bit cryptic to me. Could you elaborate? > > Thanks, > Steve > > On Jul 25, 2005, at 12:32 PM, Sean Luke wrote: > >> The simpler solution is to set NaN and infinity to very bad function >> sets. They tend to get weeded out rapidly. >>