In order for that to work you must have an evolved robot without junk DNA, because coding the same thing takes many more commands in evolved bots than in manually-coded bots.
I am evolving robots right now shvartz, using a take from the way you evolve bots , with a few mods... ITS ALL IN THE "shvartz help me out thread" (witch is kinda dead right now exsept for my occasional posts sometimes)
That reminds me , I resantly changed the system to be more user frandly, now it favors bots that take up less cpu time, Before it favored more cpu time because the timer was not synchronized with the simulation , and this a reword timer so imagine what was going on with with slower cycles. Although it is still not sensitive to slow down dew to high populations - thats the beauty of the system:
' calls main form status bar update , bau sayes: also tweacks auto save to work at the rate of the simulation favoring faster executing robots
Public Sub cyccaption(ByVal num As Single)
Dim enterval As Long
If num > 0 Then
enterval = 30000 / num * 40 / totnvegs '<------ :D
If enterval > 65535 Then enterval = 65535
Timer2.Interval = enterval 'Auto Save , Auto Reword timer
End If
MDIForm1.infos num, totalrobots, totnvegs, totvegs, SimOpts.TotBorn, SimOpts.TotRunCycle, SimOpts.TotRunTime
End Sub
The qustion indeed is: How do (we) define junk dna , thats one of the reasons I "invented" it to check by gene and not by the whole dna, here is a few examples:
cond
5 .eye5 >
5 .eye5 > 'duplicate still costs
start
'...
'or
cond
start
5 .up store
stop
cond
start
5 .aimdx store
stop
'instead of
cond
start
5 .up store
5 .aimdx store
stop
So I desided to scrap doing any calculations for gene definitions at all. (Hence , it works per gene and only takes the code after start)
Anyway Num is right currently this thread is uselss. The only reason I posted it in the "off topic" board is incase I or someone will need it in the future. (I know I needed it at one point and did not know how to properly write the equation)
