Slime mold dynamical system model

[emackev]

A demo notebook to:

• Write a (simple 1D) dynamical systems model of slime mold
• Simulate the slime mold model
  • from known parameters
  • from a Bayesian model with unknown parameters and measurement noise
• Use ChiRho to infer the model parameters from data
• Use ChiRho to infer what would happen if we intervene

[SamWitty]

@emackev , are you still interested in turning this notebook into a full-blown example like https://basisresearch.github.io/chirho/backdoor.html or https://basisresearch.github.io/chirho/mediation.html?

[emackev]

@emackev , are you still interested in turning this notebook into a full-blown example like https://basisresearch.github.io/chirho/backdoor.html or https://basisresearch.github.io/chirho/mediation.html?

Yes! What's the best way to go about this? Should it be a stand-along thing, or combined with another demo?

[SamWitty]

I think stand-alone thing is best, to keep the size of each tutorial relatively small. I'll rearrange the dynamical systems tutorials into a coherent narrative together after we have them all in place.

It seems like the main thing that's missing is: (1) explanatory text, (2) a "success" story (it's possible it's already a success, we'd just need to explore the posterior predictive distributions), and (3) ideally real or real-ish data.