configuration_sparse_multiplayers module

Configuration for the simulations, for the multi-players case with sparse activated players.

configuration_sparse_multiplayers.HORIZON = 10000

HORIZON : number of time steps of the experiments. Warning Should be >= 10000 to be interesting “asymptotically”.

configuration_sparse_multiplayers.REPETITIONS = 4

REPETITIONS : number of repetitions of the experiments. Warning: Should be >= 10 to be statistically trustworthy.

configuration_sparse_multiplayers.DO_PARALLEL = True

To profile the code, turn down parallel computing

configuration_sparse_multiplayers.N_JOBS = -1

Number of jobs to use for the parallel computations. -1 means all the CPU cores, 1 means no parallelization.

configuration_sparse_multiplayers.NB_PLAYERS = 2

NB_PLAYERS : number of players for the game. Should be >= 2 and <= number of arms.

configuration_sparse_multiplayers.ACTIVATION = 1.0

ACTIVATION : common probability of activation.

configuration_sparse_multiplayers.ACTIVATIONS = (1.0, 1.0)

ACTIVATIONS : probability of activation of each player.

configuration_sparse_multiplayers.VARIANCE = 0.05

Variance of Gaussian arms

configuration_sparse_multiplayers.NB_ARMS = 2

Number of arms for non-hard-coded problems (Bayesian problems)

configuration_sparse_multiplayers.ARM_TYPE

alias of Arms.Bernoulli.Bernoulli

configuration_sparse_multiplayers.MEANS = [0.3333333333333333, 0.6666666666666667]

Means of the arms

configuration_sparse_multiplayers.configuration = {'activations': (1.0, 1.0), 'averageOn': 0.001, 'environment': [{'arm_type': <class 'Arms.Bernoulli.Bernoulli'>, 'params': [0.3333333333333333, 0.6666666666666667]}], 'finalRanksOnAverage': True, 'horizon': 10000, 'n_jobs': -1, 'players': [Selfish(UCB), Selfish(UCB)], 'repetitions': 4, 'successive_players': [[Selfish(U(1..2)), Selfish(U(1..2))], [Selfish(UCB), Selfish(UCB)], [Selfish(Thompson Sampling), Selfish(Thompson Sampling)], [Selfish(kl-UCB), Selfish(kl-UCB)], [Selfish(Exp3++), Selfish(Exp3++)]], 'verbosity': 6}

This dictionary configures the experiments

configuration_sparse_multiplayers.nbArms = 2

Number of arms in the first environment