Sean,

Good point -- I will definitely get to that in part 4. Part 3 will be  
a brief explanation of some of the base classes in ECJ, and part 4  
will revisit the tutorial 1 params file, at which point the viewer  
should have enough information to understand the parameters having to  
do with the probabilities :)

Part 2 was really just to get the thing to run and play a little.

--Rob

On Dec 6, 2007, at 11:03 AM, Sean Luke wrote:

> On Dec 6, 2007, at 8:31 AM, Robert Baruch wrote:
>
>> Second part is up: How to run ECJ in Eclipse :)
>>
>> http://www.youtube.com/watch?v=CYoxqL0_m1E
>
> Had some trouble with tutorial1 at a bigger size genome?  Increases  
> in just pop size and generations won't help you: you're mutating too  
> much to converge. Try to stick to around 1/L, and bump the pop size:
>
> java ec.Evolve -file tutorial1.params -p pop.subpop.0.species.genome- 
> size=200 -p pop.subpop.0.species.mutation-prob=0.005 -p pop.subpop. 
> 0.size=70
>
> Sean
>