Skip to content

Latest commit

 

History

History
164 lines (148 loc) · 9.8 KB

CHANGELOG.md

File metadata and controls

164 lines (148 loc) · 9.8 KB
Raw

Changelog

2024/10/14 - 00 - v0.0.8 - Instantiation and GitHub CI

  • Instantiated backends and solvers with different numerical libraries:
    • Added context_manager modules to quantrl.backends and quantrl.solvers packages.
    • Added support for JIT-compilation in quantrl.backends package.
    • Changed naming convention for GPU-based options as gpu from cuda.
  • Added tox-based testing and continuous integration with GitHub workflows.
  • Updated all modules with PEP-based code styling and added pylintrc.
  • Updated requirements.txt and added version to quantrl.__init__.
  • Replaced setup.py with pyproject.toml.
  • Updated docs and README.

2024/08/09 - 00 - v0.0.7 - Update Callback

  • Fixed issue with cache indexing in quantrl.envs.base module.
  • Updated best reward callback in quantrl.utils module.
  • Added seaborn to requirements and setup.

2027/07/23 - 00 - v0.0.7 - Fix Learning Curve Plotter

  • Fixed issue with learning curve data in quantrl.envs.base module.
  • Fixed averaging of reward data in quantrl.plotters.LearningCurvePlotter class.

2027/07/22 - 00 - v0.0.7 - Fix Seed and Add Plotter

  • Added entropy-based seed sequencing for generators in quantrl.backends package.
  • Minor fixes to quantrl.envs.base module
  • Added option to save and load files directly using quantrl.io.FileIO class.
  • Added quantrl.plotters.LearningCurvePlotter class to plot learning curves directly.
  • Added option for averaging and saving in quantrl.utils.SaveOnBestMeanRewardCallback class.
  • Updated docs and CONTRIBUTING.

2024/07/06 - 00 - v0.0.7 - Add Callback Util

  • Fixed plot buffer update in quantrl.envs.base module.
  • Added garbage collection in quantrl.io.FileIO class.
  • Support for more plots in quantrl.plotters.TrajectoryPlotter class.
  • Added quantrl.utils module for utility functions.
  • Updated README, requirements and setup.

2024/05/29 - 00 - v0.0.7 - Added Documentation

  • Added documentation for all modules.
  • Added uniform, norm, if_else and argmax methods in quantrl.backends package.
  • Updated quantrl.envs package:
    • Update default parameters and added option to buffer plots in base.BaseEnv class.
    • Added option to cache only the properties in base.BaseEnv class.
    • Minor fixes to stochastic.LinearEnv class.
  • Added complex datatype option to quantrl.solvers package.
  • Added CODE_OF_CONDUCT, CONTRIBUTING and LICENSE.

2024/04/23 - 00 - v0.0.7 - Hotfix for PRNGs

  • Updated generator methods in quantrl.backends package.
  • Updated quantrl.envs.base module:
    • Added seed parameter to BaseEnv class.
    • Initialized states, observations, properties and results, and made removed environment validation during initialization of BaseGymEnv and BaseSB3Env classes.
    • Minor updates to cache and plot in step and evolve methods of BaseGymEnv class.
  • Minor fixes to quantrl.envs.stochastic module.

2024/04/04 - 00 - v0.0.7 - Added Observation Noise

  • Added option to add standard deviations of the observed states Observations from the actual states States through the observation_stds parameter in the reset and update methods.
  • Renamed reset_observations method to reset_states to reset the actual states States.
  • Renamed _update_observations method to _update_states to update the actual states States.
  • Updated documentation and README.

2024/03/23 - 00 - v0.0.7 - Updated Stochastic Environment

  • Added iterate_i method to support single loops in quantrl.backends package.
  • Updated quantrl.envs.stochastic.LinearEnv with support for JAX loops.
  • Removed iterative solvers from quantrl.solvers package.
  • Updated README and setup.

2024/03/22 - 00 - v0.0.7 - Updated FileIO

  • Updated quantrl.backends package:
    • Renamed occurences of tensor to typed and array to numpy.
    • Renamed lib attribute to library for the backend library.
    • Added numpy option to dtype_from_str method.
    • Added is_typed, generator and integers methods.
  • Updated quantrl.envs package:
    • Added data_idxs option for selective storage.
    • Renamed disk_cache_size to cache_dump_interval.
    • Removed preserve_dtype option. All backend data-types are now preserved.
    • Updated default value of observation_space_range to [-1e12, 1e12] and added check_truncation method to check if observations are out of bounds..
    • Added t_ssz attribute for the actual step size.
    • Added dedicated variables for caching, saving and plotting in vectorized environments.
  • Updated quantrl.io.FileIO class:
    • Renamed disk_cache_size to cache_dump_interval.
    • Removed option to update cache in parts and added dump_part method to directly dump part data of batches.
    • Added threading to support asynchronous cache dump.
  • Templated iterative solvers to quantrl.solvers package.
  • Updated README and setup.

2024/03/18 - 00 - v0.0.6 - PyTorch and JAX Backends

  • Added quantrl.backends package:
    • Added abstract base class BaseBackend for different backends in base module.
    • Interfaced JAX, PyTorch and NumPy backends in jax, torch and numpy modules.
  • Updated quantrl.envs.base module:
    • Created and documented abstract methods for BaseEnv class.
    • Renamed _step method to _update_observations.
    • Interfaced backends in BaseEnv, BaseGymEnv and BaseSB3Env classes.
    • Added parameter preserve_dtype to preserve the interfaced tensor data-type in BaseEnv class.
    • Removed truncation information from step_async in BaseSB3Env class.
  • Updated quantrl.envs.deterministic module:
    • Interfaced backends in LinearizedHOEnv and LinearizedHOVecEnv classes.
    • Added template methods for get_A, get_D and get_mode_rates and updated documentation.
  • Interfaced backends and removed MCQT support in quantrl.envs.stochastic.LinearEnv class.
  • Added base solver module and interfaces for three different IVP solvers (jax.DiffraxIVPSolver, torch.TorchDiffEqIVPSolver and numpy.SciPyIVPSolver classes) in quantrl.solvers package.
  • Updated README and setup.

2024/03/01 - 00 - v0.0.5 - Vectorized Environments

  • Updated quantrl.envs.base module:
    • reset_Observations, get_Properties and get_Reward methods of all classes are renamed to reset_observations, get_properties and get_reward, with reset_observations returning the initial observations.
    • Removed n_trajectories parameter and added n_properties parameter in all classes.
    • Added file_prefix parameter and action_steps attribute to BaseEnv class.
    • Renamed file_prefix attribute to file_path_prefix for the complete path.
    • Updated the workflow of step methods of all classes.
    • plot_learning_curve method of BaseEnv class takes n_episodes parameters for running average.
    • replay_trajectories and close methods of BaseEnv class require n_episodes parameter.
    • close method of BaseEnv class takes an additional save_replay parameter.
    • Added save parameter to close method of BaseGymEnv class.
    • Added close parameter to evolve method of BaseGymEnv class.
    • Added BaseSB3Env class to support Stable-Baselines3 vectorized environments.
  • Updated quantrl.envs.deterministic module:
    • Removed n_trajectories and added n_properties parameters to LinearizedHOEnv class.
    • Removed actions parameter from LinearizedHOEnv._step method.
    • Added LinearizedHOVecEnv class to interface vectorized environments.
  • Updated quantrl.envs.stochastic.LinearEnv class:
    • Removed n_trajectories and added n_properties parameters.
    • Removed actions parameter from _step and _step_wiener methods.
  • Updated quantrl.io.FileIO class:
    • Added option to initialize with data_shape parameter.
    • Renamed max_cache_size parameter to disk_cache_size.
    • Added dump_cache parameter in close method.
  • Added option to save plots in save_dir directory in quantrl.plotters.TrajectoryPlotter class.
  • Updated README, requirements and setup.

2024/02/20 - 00 - v0.0.4 - Implemented kwargs

  • Updated quantrl.envs.base.BaseGymEnv class:
    • Implemented keyword arguments in initialization.
    • Properties are saved if the child method exists.
  • Minor fixes to quantrl.envs.deterministic.LinearizedHOEnv class.
  • Updated Wiener process for drift matrix in quantrl.envs.stochastic.LinearEnv class.
  • Reversed the sequence of arguments in quantrl.solvers.differential.SciPyIVPSolver class.
  • Update tick options in quantrl.plotter module.
  • Updated README and setup.

2024/02/10 - 00 - v0.0.3 - Added Wiener Processes

  • Restructured quantrl.envs.rl module:
    • Created quantrl.envs.base module with BaseGymEnv.
    • Created quantrl.envs.deterministic module with LinearizedHOEnv.
    • Created quantrl.envs.stochastic module with HOEnv.
  • Buffer fixes to quantrl.solvers.differential.SciPyIVPSolver.
  • Fixed cache dump and underflow in quantrl.io.FileIO.
  • Added options to make gifs and save plots in quantrl.plotters.TrajectoryPlotter.
  • Updated README and setup.

2024/01/16 - 00 - v0.0.2 - DDE Support

  • Replaced quantrl.envs.BaseEnv with quantrl.envs.rl.
  • Added quantrl.solvers.differential module with DDE support.
  • Renamed quantrl.solvers.QCMSolver to quantrl.solvers.measure.
  • Replaced quantrl.log with quantrl.io.
  • Replaced quantrl.utils with quantrl.plotters.
  • Added README, requirements and setup.

2023/05/20 - 00 - v0.0.1 - Initial Commit

  • Added quantrl.envs.BaseEnv module.
  • Added quantrl.solvers.QCMSolver module.
  • Added quantrl.log and quantrl.utils module.
  • Added CHANGELOG, README and setup.