7 Commits

Author SHA1 Message Date
maxhuettenrauch
ade85ab32b
Feature/algo eval (#1074)
# Changes

## Dependencies

- New extra "eval"

## Api Extension
- `Experiment` and `ExperimentConfig` now have a `name`, that can
however be overridden when `Experiment.run()` is called
- When building an `Experiment` from an `ExperimentConfig`, the user has
the option to add info about seeds to the name.
- New method in `ExperimentConfig` called
`build_default_seeded_experiments`
- `SamplingConfig` has an explicit training seed, `test_seed` is
inferred.
- New `evaluation` package for repeating the same experiment with
multiple seeds and aggregating the results (important extension!).
Currently in alpha state.
- Loggers can now restore the logged data into python by using the new
`restore_logged_data`

## Breaking Changes
- `AtariEnvFactory` (in examples) now receives explicit train and test
seeds
- `EnvFactoryRegistered` now requires an explicit `test_seed`
- `BaseLogger.prepare_dict_for_logging` is now abstract

---------

Co-authored-by: Maximilian Huettenrauch <m.huettenrauch@appliedai.de>
Co-authored-by: Michael Panchenko <m.panchenko@appliedai.de>
Co-authored-by: Michael Panchenko <35432522+MischaPanch@users.noreply.github.com>
2024-04-20 23:25:33 +00:00
Daniel Plop
eb0215cf76
Refactoring/mypy issues test (#1017)
Improves typing in examples and tests, towards mypy passing there.

Introduces the SpaceInfo utility
2024-02-06 14:24:30 +01:00
Dominik Jain
05a8cf4e74 Refactoring, improving class name EnvFactoryGymnasium -> EnvFactoryRegistered 2024-01-16 14:52:31 +01:00
Dominik Jain
eaab7b0a4b Improve environment factory abstractions in high-level API:
* EnvFactory now uses the creation of a single environment as
   the basic functionality which the more high-level functions build
   upon
 * Introduce enum EnvMode to indicate the purpose for which an env
   is created, allowing the factory creation process to change its
   behaviour accordingly
 * Add EnvFactoryGymnasium to provide direct support for envs that
   can be created via gymnasium.make
     - EnvPool is supported via an injectible EnvPoolFactory
     - Existing EnvFactory implementations are now derived from
       EnvFactoryGymnasium
 * Use a separate environment (which uses new EnvMode.WATCH) for
   watching agent performance after training (instead of using test
   environments, which the user may want to configure differently)
2024-01-12 17:13:42 +01:00
Dominik Jain
b5a891557f Revert to simplified environment factory, removing unnecessary config object
(configuration shall be part of the factory instance)
2023-10-24 13:14:23 +02:00
Dominik Jain
6cbee188b8 Change interface of EnvFactory to ensure that configuration
of number of environments in SamplingConfig is used
(values are now passed to factory method)

This is clearer and removes the need to pass otherwise
unnecessary configuration to environment factories at
construction
2023-10-19 11:37:20 +02:00
Dominik Jain
50ac385321 Add some basic tests for high-level experiment builder API 2023-10-18 20:44:16 +02:00