Rcognita: A framework for DP and RL algorithm development, testing, and simulation

Welocme to rcognita’s documentation! This wiki contains all the nescessary information to get started with usage or development. Follow the installation instructions to get rcognita or check out rcognita’s gitflic page if you’d like to browse releases or fork the repository.

Modules

Classes

ComplementedConfig(cfg[, config_path])

A config object, generated by rcognita’s config pipeline.

main(*args[, logger, callbacks])

The decorator used to invoke rcognita’s config pipeline.

rcognita.actors

This module contains actors, i.e., entities that directly calculate actions.

rcognita.callbacks

This module contains callbacks.

rcognita.controllers

This module contains high-level structures of controllers (agents).

rcognita.critics

This module containing critics, which are integrated in controllers (agents).

rcognita.loggers

This module contains the logger interface along with concrete realizations for each separate system.

rcognita.models

This module contains model classes.

rcognita.objectives

Module that contains general objectives functions that can be used by various entities of the framework.

rcognita.observers

rcognita.optimizers

This module contains optimization routines to be used in optimal controllers, actors, critics etc.

rcognita.predictors

Module that contains state or observation (depending on the context) predictors.

rcognita.scenarios

This module contains various simulation scenarios.

rcognita.simulator

This module contains one single class that simulates controller-system (agent-environment) loops.

rcognita.solvers

rcognita.systems

This module contains a generic interface for systems (environments) as well as concrete systems as realizations of the former

rcognita.visualization