configuration_multiplayers_with_aggregation module¶
Configuration for the simulations, for multi-players with aggregation.
-
configuration_multiplayers_with_aggregation.
HORIZON
= 10000¶ HORIZON : number of time steps of the experiments. Warning Should be >= 10000 to be interesting “asymptotically”.
-
configuration_multiplayers_with_aggregation.
REPETITIONS
= 200¶ REPETITIONS : number of repetitions of the experiments. Warning: Should be >= 10 to be statistically trustworthy.
-
configuration_multiplayers_with_aggregation.
DO_PARALLEL
= True¶ To profile the code, turn down parallel computing
-
configuration_multiplayers_with_aggregation.
N_JOBS
= -1¶ Number of jobs to use for the parallel computations. -1 means all the CPU cores, 1 means no parallelization.
-
configuration_multiplayers_with_aggregation.
NB_PLAYERS
= 6¶ NB_PLAYERS : number of players for the game. Should be >= 2 and <= number of arms.
-
configuration_multiplayers_with_aggregation.
collisionModel
(t, arms, players, choices, rewards, pulls, collisions)¶ The best collision model: none of the colliding users get any reward
-
configuration_multiplayers_with_aggregation.
VARIANCE
= 0.05¶ Variance of Gaussian arms
-
configuration_multiplayers_with_aggregation.
CACHE_REWARDS
= False¶ Should we cache rewards? The random rewards will be the same for all the REPETITIONS simulations for each algorithms.
-
configuration_multiplayers_with_aggregation.
NB_ARMS
= 12¶ Number of arms for non-hard-coded problems (Bayesian problems)
-
configuration_multiplayers_with_aggregation.
LOWER
= 0.0¶ Default value for the lower value of means
-
configuration_multiplayers_with_aggregation.
AMPLITUDE
= 1.0¶ Default value for the amplitude value of means
-
configuration_multiplayers_with_aggregation.
ARM_TYPE
¶ alias of
Arms.Bernoulli.Bernoulli
-
configuration_multiplayers_with_aggregation.
ENVIRONMENT_BAYESIAN
= False¶ True to use bayesian problem
-
configuration_multiplayers_with_aggregation.
MEANS
= [0.05, 0.1318181818181818, 0.21363636363636362, 0.2954545454545454, 0.3772727272727272, 0.459090909090909, 0.5409090909090909, 0.6227272727272727, 0.7045454545454545, 0.7863636363636363, 0.868181818181818, 0.95]¶ Means of arms for non-hard-coded problems (non Bayesian)
-
configuration_multiplayers_with_aggregation.
configuration
= {'averageOn': 0.001, 'change_labels': {0: 'Aggr(rhoRand, RandTopM, MCTopM, Selfish) kl-UCB', 1: 'rhoRand kl-UCB', 2: 'RandTopM kl-UCB', 3: 'MCTopM kl-UCB', 4: 'Selfish kl-UCB', 5: 'Centralized kl-UCB'}, 'collisionModel': <function onlyUniqUserGetsReward>, 'environment': [{'arm_type': <class 'Arms.Bernoulli.Bernoulli'>, 'params': [0.05, 0.1318181818181818, 0.21363636363636362, 0.2954545454545454, 0.3772727272727272, 0.459090909090909, 0.5409090909090909, 0.6227272727272727, 0.7045454545454545, 0.7863636363636363, 0.868181818181818, 0.95]}], 'finalRanksOnAverage': True, 'horizon': 10000, 'n_jobs': -1, 'players': [Selfish(UCB), Selfish(UCB), Selfish(UCB), Selfish(UCB), Selfish(UCB), Selfish(UCB)], 'plot_lowerbounds': False, 'repetitions': 200, 'successive_players': [[<Policies.Aggregator.Aggregator object>, <Policies.Aggregator.Aggregator object>, <Policies.Aggregator.Aggregator object>, <Policies.Aggregator.Aggregator object>, <Policies.Aggregator.Aggregator object>, <Policies.Aggregator.Aggregator object>], [rhoRand(Aggregator($N=3$)), rhoRand(Aggregator($N=3$)), rhoRand(Aggregator($N=3$)), rhoRand(Aggregator($N=3$)), rhoRand(Aggregator($N=3$)), rhoRand(Aggregator($N=3$))], [RandTopM(Aggregator($N=3$)), RandTopM(Aggregator($N=3$)), RandTopM(Aggregator($N=3$)), RandTopM(Aggregator($N=3$)), RandTopM(Aggregator($N=3$)), RandTopM(Aggregator($N=3$))], [MCTopM(Aggregator($N=3$)), MCTopM(Aggregator($N=3$)), MCTopM(Aggregator($N=3$)), MCTopM(Aggregator($N=3$)), MCTopM(Aggregator($N=3$)), MCTopM(Aggregator($N=3$))], [Selfish(Aggregator($N=3$)), Selfish(Aggregator($N=3$)), Selfish(Aggregator($N=3$)), Selfish(Aggregator($N=3$)), Selfish(Aggregator($N=3$)), Selfish(Aggregator($N=3$))], [CentralizedMultiplePlay(Aggregator($N=3$)), CentralizedMultiplePlay(Aggregator($N=3$)), CentralizedMultiplePlay(Aggregator($N=3$)), CentralizedMultiplePlay(Aggregator($N=3$)), CentralizedMultiplePlay(Aggregator($N=3$)), CentralizedMultiplePlay(Aggregator($N=3$))]], 'verbosity': 6}¶ This dictionary configures the experiments
-
configuration_multiplayers_with_aggregation.
nbArms
= 12¶ Number of arms in the first environment