Agents.jl: Dynamic ABM plotting

[Datseris]

It would be great if we could support dynamic plotting where one can interactively see how the model evolves. Mesa, the Python ABM framework has some support for that: https://mesa.readthedocs.io/en/master/tutorials/adv_tutorial.html

A good candidate which has a lot of support for interactive visualizations is Makie.jl, because it hooks up very well with Observables.jl . Other alternatives would probably be packages from JuliaGizmos for a web-based application.

Any other suggestions/alternatives on how this could be done are welcome.

(notice that this functionality will reside in AgentsPlots.jl)

[itsdfish]

@Datseris, I have limited experience with package development and visualization, but I will see if I can create a working prototype with Makie until a more permanent solution can be made. Would you recommend passing an optional plot_step! function to step!, which would be called after all agents are updated?

[Datseris]

Hm, I think the best option would be to create an independent plotting function whose inputs are abm, agent_step!, model_step! as well as all other arguments relevant for plotting. step! is already quite heavy in functionality and thus I would try to keep plotting separated from step!. Thanks for your interest by the way!

[Libbum]

Can we consider JuliaDynamics/InteractiveChaos.jl#7 to have championed this issue and close it?