Print

Print


Dear,
 
I've done as instructed, but when I use the output log file as the input
file for this parameter: pop.subpop.0.file = <filename>, I got the
error:
 
| ECJ
| An evolutionary computation system (version 11)
| Copyright 2004 by Sean Luke
| URL: http://www.cs.umd.edu/projects/plus/ec/ecj/
| Mail: [log in to unmask]
| Date: March 5, 2004
| Current Java: 1.4.2_04 / Java HotSpot(TM) Client VM-1.4.2_04-b05
| Suggested Minimum Java: 1.2.2
| Required Minimum Java: 1.1
 
 
Threads:  breed/1 eval/1
Seed: 4357
Setting up
Initializing Generation 0
FATAL ERROR:
Reading Line 2: Bad 'Evaluated?' line.
 
If you have a sample input file for initial population, please send me.
 
Thank you,
Duong.
 
-----Original Message-----
From: ECJ Evolutionary Computation Toolkit
[mailto:[log in to unmask]] On Behalf Of Sean Luke
Sent: Saturday, July 24, 2004 2:18 AM
To: [log in to unmask]
Subject: Re: Customized initial population and execution time in ECJ
 
On Jul 23, 2004, at 1:40 PM, #TA NGUYEN BINH DUONG# wrote:
 
> Thank you for your help. But what's the format of the file to init the
> population? I could not find it anywhere in the documents provided.
 
Each subpopulation can load from a separate file, where each individual
is written, in order, to that file using the format produced with
Individual.printIndividual(...).
 
If you'd like to see what such a file looks like, change SimpleFinisher
to look like this, and it will write out a file at the end of the
evolutionary run for Subpopulation 0:
 
package ec.simple;
import ec.Finisher;
import ec.EvolutionState;
import ec.util.Parameter;
 
public class SimpleFinisher extends Finisher
     {
     public void setup(final EvolutionState state, final Parameter base)
{ }
     public void finishPopulation(final EvolutionState state, final int
result)
         {
         try
             {
             int log = state.output.addLog(
                 new java.io.File("example.log"),0,false,false);
             for (int
x=0;x<state.population.subpops[0].individuals.length;x++)
                 {
                 ec.Individual i =
state.population.subpops[0].individuals[x];
                 i.printIndividual(state,log,100);
                 }
             }
         catch (java.io.IOException e) { }
         return;
         }
     }