Strategy Base

Base class containing the core methods of CRLD agents in strategy space

strategybase

 strategybase (env, learning_rates:Union[float,Iterable],
               discount_factors:Union[float,Iterable],
               choice_intensities:Union[float,Iterable]=1.0,
               use_prefactor=False, opteinsum=True, **kwargs)

Base class for deterministic strategy-average independent (multi-agent) temporal-difference reinforcement learning in strategy space.

	Type	Default	Details
env			An environment object
learning_rates	Union		agents’ learning rates
discount_factors	Union		agents’ discount factors
choice_intensities	Union	1.0	agents’ choice intensities
use_prefactor	bool	False	use the 1-DiscountFactor prefactor
opteinsum	bool	True	optimize einsum functions
kwargs

Further optional paramerater inherting from abase:

	Type	Default	Details
use_prefactor	bool	False	use the 1-DiscountFactor prefactor
opteinsum	bool	True	optimize einsum functions

source

strategybase.step

 strategybase.step (Xisa)

Performs a learning step along the reward-prediction/temporal-difference error in strategy space, given joint strategy Xisa.

	Type	Details
Xisa		Joint strategy
Returns	tuple	(Updated joint strategy, Prediction error)

source

strategybase.reverse_step

 strategybase.reverse_step (Xisa)

*Performs a reverse learning step in strategy space, given joint strategy Xisa.

This is useful to compute the separatrix of a multistable regime.*

	Type	Details
Xisa		Joint strategy
Returns	tuple	(Updated joint strategy, Prediction error)

source

strategybase.zero_intelligence_strategy

 strategybase.zero_intelligence_strategy ()

Returns strategy Xisa with equal action probabilities.

source

strategybase.random_softmax_strategy

 strategybase.random_softmax_strategy ()

Returns softmax strategy Xisa with random action probabilities.

source

strategybase.id

 strategybase.id ()

Returns an identifier to handle simulation runs.