Road map

Implementation-related tasks

Odds & ends (clean up original model)

1. Remove dead agents from graph entirely (currently they stay as zombie neighbors)
2. Permit more than two agents per proto (currently exactly two per proto)

(Possible) model features to implement

1. Proto resource pooling (insurance)
   • Each iteration that its personal wealth is over threshold X, an agent contributes a fixed amount to the proto's account.
   • Each iteration that its personal wealth is below threshold Y (<X), an agent withdraws enough from the proto's account to reach Y.
   • Q: If proto account reaches zero...?
2. Proto dissolution
   • Q: Under what conditions? (Or is this the same as the previous item?)
3. Proto resource pooling (investment)
   • Q: how would this work?
4. Limited window of opportunity to join protos (inspired by Handbook ch.6)
   • Q: how to model the limited opportunity window?
     • A maximum number of agents per proto (which is same as current model, actually, if max=2)?
     • A length of time during which a proto is "joinable," after which it accepts no new members?
5. Other network models
6. Some agents inherently do, some do not, form protos
   • Exogenous boolean attribute of agents, randomly determined at creation time

Analysis features to implement

1. Wealth distro
   1. Histogram of agent wealth (snapshot at end of simulation)
   2. Plot of Gini coefficient over time
   3. Graph assortativity by wealth attribute
   4. Q: others???
2. Proto membership
   1. Plot of number-of-protos over time
   2. Histogram of proto size (snapshot at end of simulation)
   3. Q: others???
3. Parameter sweeps
   1. Erdos-Renyi λ parameter