Group Selection

We include Group Selection for a population with a fixed number of groups m consisting of n individuals each. See the white paper (link on the wiki homepage) for the exact implementation. Currently both the WrightFisher and Moran dynamics can be used for processes with group selection. The parameters of interest are,

• rate : This is the probability that a group-level event occurs at each time-step.
• number_groups: Number of groups in the simulation.
• pop_size: This is the total population size in the system. It gets divided equally among the m groups.
• selection_strengthI: This is the selection strength used at the individual level while mapping payoffs to fitness.
• selection_strengthG: This is the selection strength used at the group level while mapping the average payoffs of groups to their average group fitness.

These parameters are arguments of the DynamicsSimulator class and can be specified when creating objects of the GameDynamicsWrapper and VariedGame class. For example,

s = GameDynamicsWrapper(PrisonersDilemma,WrightFisher,dynamics_kwargs={'number_groups':20,'rate':0.2})
s.simulate_many(num_iterations=100, num_gens=100, pop_size=100)