Limit on the number of agents?

[Norman Jacknis]

Is there a limit on the number of agents in NetLogo?  When we run it with data about 3000 agents it works OK.  When we go to 30,000 agents, the software initializes OK, but nothing happens when clicking on Go (to run the simulation).

Thanks, Norm

[Michael Tamillow]

One definition of computer science is the study of the limits of computation. Consider the computational complexity of your program.

30,000 agents doing what?

Netlogo has significant overhead, particularly rendering. If you run headless, without the Ui for just data collection, it can handle a greater workload.

Sent from my iPhone

On Apr 17, 2024, at 6:18 PM, 'Norman Jacknis' via netlogo-users <netlog...@googlegroups.com> wrote:



Is there a limit on the number of agents in NetLogo?  When we run it with data about 3000 agents it works OK.  When we go to 30,000 agents, the software initializes OK, but nothing happens when clicking on Go (to run the simulation).

Thanks, Norm

--
You received this message because you are subscribed to the Google Groups "netlogo-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to netlogo-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/netlogo-users/12f84cca-69b6-4aa9-8592-916a5d2c63d3n%40googlegroups.com.

[Aaron Andre Brandes]

Hi Norm,

There is no fixed limit. However each agent (turtle, patch, link) and each variable take up memory. For any given amount of memory performance will become exceedingly slow if you have too many agents. You may be able to dedicate more memory to NetLogo.

See https://ccl.northwestern.edu/netlogo/docs/faq.html#how-big-can-my-model-be-how-many-turtles-patches-procedures-buttons-and-so-on-can-my-model-contain

• Aaron

 

-- 

Aaron Brandes, Software Developer

Center for Connected Learning and Computer-Based Modeling

 

 

--

[Steve Railsback]

We routinely run models with > 1,000,000 turtles. There are many reasons why performance can slow rapidly with the number of agents, and they can almost always be avoided. I strongly suggest you consult this paper to determine why your model becomes slow:

http://jasss.soc.surrey.ac.uk/20/1/3.html