bnelearn.experiment.single_item_experiment module

This module implements Experiments on single items

class bnelearn.experiment.single_item_experiment.AffiliatedObservationsExperiment(config: ExperimentConfig)[source]

Bases: SingleItemExperiment

A Single Item Experiment that has the same valuation prior for all participating bidders. For risk-neutral agents, a unique BNE is known.

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.ContestExperiment(config: ExperimentConfig)[source]

Bases: SymmetricPriorSingleItemExperiment

This class implements a symmetric Contest Experiment

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
pretrain_transform(player_pos: Optional[int] = None) → callable[source]

Some experiments need specific pretraining transformations. In most cases, pretraining to the truthful bid (i.e. the identity function) is sufficient.

Args:
player_position (:int:) the player for which the transformation is

requested.

Returns

(:callable:) pretraining transformation

sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.GaussianSymmetricPriorSingleItemExperiment(config: ExperimentConfig)[source]

Bases: SymmetricPriorSingleItemExperiment

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.MineralRightsExperiment(config: ExperimentConfig)[source]

Bases: SingleItemExperiment

A Single Item Experiment that has the same valuation prior for all participating bidders. For risk-neutral agents, a unique BNE is known.

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.SingleItemExperiment(config: ExperimentConfig)[source]

Bases: Experiment, ABC

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
static get_risk_profile(risk) → str[source]

Used for logging and checking existence of bne

input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.SymmetricPriorSingleItemExperiment(config: ExperimentConfig)[source]

Bases: SingleItemExperiment

A Single Item Experiment that has the same valuation prior for all participating bidders. For risk-neutral agents, a unique BNE is known.

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.TwoPlayerAsymmetricBetaPriorSingleItemExperiment(config: ExperimentConfig)[source]

Bases: SingleItemExperiment

A single item experiment where two bidders have different beta priors.

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.TwoPlayerAsymmetricUniformPriorSingleItemExperiment(config: ExperimentConfig)[source]

Bases: SingleItemExperiment

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int
class bnelearn.experiment.single_item_experiment.UniformSymmetricPriorSingleItemExperiment(config: ExperimentConfig)[source]

Bases: SymmetricPriorSingleItemExperiment

action_size: int
b_opt: torch.Tensor
bidders: Iterable[Bidder]
bne_env: AuctionEnvironment
bne_utilities: Tensor
env: Environment
epoch: int
input_length: int
learners: Iterable[learners.Learner]
mechanism: Mechanism
models: Iterable[torch.nn.Module]
n_models: int
observation_size: int
plot_xmax: float
plot_xmin: float
plot_ymax: float
plot_ymin: float
positive_output_point: Tensor
sampler: ValuationObservationSampler
v_opt: torch.Tensor
valuation_size: int