coba.environments

Environment creators and modifiers.

Friendly API

Environments

A friendly API for common environment functionality.

Sources

CsvSource	Load a csv dataset.
ArffSource	Load an arff dataset.
LibSvmSource	Load a libsvm dataset.
ManikSource	Load a manik dataset.
OpenmlSource	Load an openml.org dataset.

Simulations

LambdaSimulation	A contextual bandit environment created from generative lambda functions.
SupervisedSimulation	A contextual bandit environment created from supervised data.
OpenmlSimulation	A supervised simulation created from an openml dataset.
LinearSyntheticSimulation	A synthetic simulation whose rewards are linear with respect to the given reward features.
NeighborsSyntheticSimulation	A synthetic simulation whose reward values are determined by neighborhoods.
KernelSyntheticSimulation	A synthetic simulation whose reward function is created from kernel basis functions.
MLPSyntheticSimulation	A synthetic simulation whose reward function belongs to the MLP family.

Filters

Batch	Batch interactions.
BatchSafe	Make a filter agnostic to batching.
Binary	Transform reward values to 1 or 0.
Cache	Cache given interactions.
Chunk	Indicate where to chunk.
Cycle	Cycle reward values.
Densify	Ensure that features are dense.
Finalize	Final preparation for built-in Evaluators.
Flatten	Flatten contexts and actions.
Grounded	Transform simulated interactions to IGL interactions.
Impute	Impute missing values (nan) in Interaction contexts.
Logged	Transform simulated interactions to logged interactions.
Noise	Add noise to values.
OpeRewards	Transform logged interactions to simulated interactions.
Repr	Change representation of categorical data.
Reservoir	Take n random interactions.
Riffle	Riffle shuffle interactions.
Scale	Scale and shift features.
Shuffle	Shuffle interaction order.
Slice	Take a slice of interactions.
Sort	Sort interactions by features.
Sparsify	Ensure that features are sparse.
Take	Take the first n interactions.
Unbatch	Unbatch interactions.
Where	Select for characteristics