Sender: |
|
Date: |
Tue, 13 Jun 2006 10:45:05 -0400 |
MIME-version: |
1.0 (Apple Message framework v750) |
Reply-To: |
|
Content-type: |
text/plain; charset=US-ASCII; delsp=yes; format=flowed |
Subject: |
|
From: |
|
In-Reply-To: |
|
Content-transfer-encoding: |
7bit |
Comments: |
|
Parts/Attachments: |
|
|
If you're doing GP, out of memory errors will be common: GP
individuals bloat unless you use one of many techniques to keep them
under control.
In this case, you need to increase the heap size. Here we increase
it to 500M:
java -Xmx500M ec.Evolve ...
If you're doing something *other* than GP, sure, it could be a leak
of some sort; though ECJ hasn't historically had many leaks -- hash
table overfills etc ...
On Jun 13, 2006, at 6:28 AM, Alberto Cuesta wrote:
> Hello all,
>
> I'm running Genetic Programming experiments, and I get easily this
> error:
>
> Exception in thread "EvolutionAgent.2508451.8329630"
> java.lang.OutOfMemoryError: Java heap space
>
> I think that it's that the population genotypes grow and exceed
> over a certain memory limit assigned to the JVM, but I could be
> wrong and have a memory leak somewhere. Has anyone some directions
> to investigate or solve this?
>
> Thanks,
> Alberto
|
|
|