gama.old icon indicating copy to clipboard operation
gama.old copied to clipboard

Published 20 hours ago • verified publisher icon gama-platform

[Batch] batch experiment from tutorial Predator Prey runs silently into errors

Open chapuisk opened this issue 3 years ago • 9 comments

Describe the bug When dealing with batch experiment on a model of the library, i.e. predator prey model 13.gaml, the execution slow drastically at some random number of replicates (optimization method). But there is no error message, but simulations get very slow and experiment never ends. I tried with git version and got an error from eclipse (which is not a Gama error) but I cannot find any good explanation to it.

To Reproduce Steps to reproduce the behavior:

  1. Go to Tutorials > Predator Prey > models > Model 13.gaml
  2. Click on Optimization
  3. See (no) error but experiment plan stuck

Expected behavior The simulation plan should run to the end or at least throw an explicit exception

Desktop (please complete the following information):

  • OS: macOS Monterey
  • GAMA version: 1.8.2RC2 and 1.8.2 git version
  • Java version: 17

Additional context I tried with several other methods involving batch experiment and runs into the same error, not being able to identify why it is happening

chapuisk avatar Sep 13 '22 14:09 chapuisk

I cannot reproduce it in the git version (between 110 and 144 simulations in ~30s each time). What error do you get ?

AlexisDrogoul avatar Sep 13 '22 14:09 AlexisDrogoul

Here is the stack trace of the error launch in eclipse (not in Gama).

Screen Shot 2022-09-14 at 14 22 41 (2)

By the way, I have been able to reproduce it in 1.8.2 RC1 and RC2, and with the git version (from which I took the capture of eclipse )

chapuisk avatar Sep 14 '22 12:09 chapuisk

Ok @AlexisDrogoul, I may have misunderstood what's going on. It seems that simulations are running, but Gama is glitching (parameter windows blinks and the display of simulations running update every n simulations or show nothing). It may not impair the experiments and simulation but outputs are broken, and eclipse have an error.

chapuisk avatar Sep 14 '22 12:09 chapuisk

I do not get this error. Can you show me your preferences related to Execution ?

AlexisDrogoul avatar Sep 14 '22 14:09 AlexisDrogoul

Here is the execution parameters of Gama for the git version; while for the RC2 version there are at default values (nothing has been changed, I even rebuild and redownload everything to do the tests)

Screen Shot 2022-09-15 at 12 14 31

chapuisk avatar Sep 15 '22 10:09 chapuisk

Hum. I can't reproduce the error event after adopting your preferences (not too different from mine btw). Anything else you might have changed in your own installation of GAMA ? Do all the batch models produce the same errors ?

AlexisDrogoul avatar Sep 16 '22 02:09 AlexisDrogoul

@ptaillandier do you observe the same behaviour ?

AlexisDrogoul avatar Sep 19 '22 08:09 AlexisDrogoul

Any hint on how to reproduce this bug ? I can't do it and I really dont see where it comes from...

AlexisDrogoul avatar Sep 26 '22 08:09 AlexisDrogoul

@AlexisDrogoul I'll try few things today to reproduce more consistently, but the fact that it happened using batch makes it quite hard to fix all conditions... if I fail, I'll close the issue

chapuisk avatar Sep 29 '22 08:09 chapuisk

I tested once and I had no errors.

Le lun. 19 sept. 2022 à 15:28, Alexis Drogoul @.***> a écrit :

@ptaillandier https://github.com/ptaillandier do you observe the same behaviour ?

— Reply to this email directly, view it on GitHub https://github.com/gama-platform/gama/issues/3486#issuecomment-1250719226, or unsubscribe https://github.com/notifications/unsubscribe-auth/AALPWHIOHDBHL2AZU7A7GEDV7AP3BANCNFSM6AAAAAAQLQBDFQ . You are receiving this because you were mentioned.Message ID: @.***>

ptaillandier avatar Oct 11 '22 07:10 ptaillandier