Sender: |
|
Date: |
Wed, 9 Mar 2005 14:07:37 -0500 |
MIME-version: |
1.0 (Apple Message framework v619.2) |
Reply-To: |
|
Content-type: |
text/plain; format=flowed; charset=US-ASCII |
Subject: |
|
From: |
|
In-Reply-To: |
|
Content-transfer-encoding: |
7bit |
Comments: |
|
Parts/Attachments: |
|
|
The best solution of a run is held by your statistics object, if you're
using one. For example, if you're using SimpleStatistics, then
state.stat.best_of_run[0] is the best individual of the run in
subpopulation 0.
Sean
On Mar 9, 2005, at 1:42 PM, Sandeep Mulgund wrote:
> Hi,
>
> I need to manage a GA run from my own program code. I found an example
> in the list archives that goes something like this:
>
> SimpleEvolutionState state = (SimpleEvolutionState)
> (Evolve.make(new String[]{"-file","mytest.params"}));
> int i;
> for (i=0; i<200; i++){
> state.go();
> if (state.debugState == state.DEBUG_END)
> break;
> }
>
> This seems to work fine. Once it's done, how can I obtain the optimal
> solution (an IntegerVectorSpecies) and use it in my own code for other
> purposes?
>
> Thanks,
>
> Sandeep
|
|
|