restarting a killed job

[Thomas Evangelidis]

Greetings,

Has anyone ever managed to restart a killed job, for instance by starting the GA from the individuals that survived from the last generation? If yes could you please give me some hints to design my program accordingly? I want to run my GA on a HPC cluster but they have 24 hours job limit.

thanks,
Thomas

[Marc-André Gardner]

Hi!

You should take a look at checkpointing : http://deap.gel.ulaval.ca/doc/dev/tutorials/advanced/checkpoint.html
It's very simple and should provide what you need.

Have fun with DEAP,

Marc-André

[Thomas Evangelidis]

Hi Marc-André,

I have a small query. When I modify the eaMuPlusLambda algorithm should I create a record of the offspring object or population+offspring. Namely should I replace the following:

if halloffame is not None:
            halloffame.update(offspring)

with this:

if halloffame is not None:
            halloffame.update(offspring)
            record = stats.compile(offspring)
            logbook.record(gen=gen, evals=len(invalid_ind), **record)

or with this?

if halloffame is not None:
            halloffame.update(offspring)
            record = stats.compile(population + offspring)
            logbook.record(gen=gen, evals=len(invalid_ind), **record)

--
You received this message because you are subscribed to a topic in the Google Groups "deap-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/deap-users/9opNM5I6chg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to deap-users+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

======================================================================

Thomas Evangelidis

PhD student

University of Athens
Faculty of Pharmacy
Department of Pharmaceutical Chemistry
Panepistimioupoli-Zografou
157 71 Athens
GREECE

email: tev...@pharm.uoa.gr

          tev...@gmail.com

website: https://sites.google.com/site/thomasevangelidishomepage/