Sinergym

Latest version: v3.6.2

Safety actively analyzes 679296 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 7 of 9

1.9.5

What's Changed
* v1.9.2 Add more usage examples in Sinergym examples folder by AlejandroCN7 in https://github.com/ugr-sail/sinergym/pull/225
* load_agent.py enhancement and evaluations fix by AlejandroCN7 in https://github.com/ugr-sail/sinergym/pull/232
* Worfklow modified to publish lite version of the container by jajimer in https://github.com/ugr-sail/sinergym/pull/228
* Introducing new buildings in Sinergym by AlejandroCN7 in https://github.com/ugr-sail/sinergym/pull/229
* Configurable csvlogger elements by Melon-PieldeSapo in https://github.com/ugr-sail/sinergym/pull/230
* LoggerWrapper Enhancement and progress.csv new field by AlejandroCN7 in https://github.com/ugr-sail/sinergym/pull/234


**Full Changelog**: https://github.com/ugr-sail/sinergym/compare/v1.9.1...v1.9.5

1.9.1

What's Changed
* Fixed DRL_battery params and rewards by manjavacas in https://github.com/jajimer/sinergym/pull/206
* Reward: Temperature(s) and power variable are required in reward function definition by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/209
* Changed evaluation callback model name to <experiment_name>.zip by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/211
* Fixed RBC5Zone with static seasonal setpoints by manjavacas in https://github.com/jajimer/sinergym/pull/214
* Updated DRL_battery.py with additional customization by manjavacas in https://github.com/jajimer/sinergym/pull/216
* Doc/issue 167 by Melon-PieldeSapo in https://github.com/jajimer/sinergym/pull/210
* Sinergym 1.9.0: Default IDF controller and Dynamic variables to control by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/212
* New issue template for improvements in Sinergym by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/220
* Curriculum learning and id name in load_agent by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/223

New Contributors
* Melon-PieldeSapo made their first contribution in https://github.com/jajimer/sinergym/pull/210

**Full Changelog**: https://github.com/jajimer/sinergym/compare/v1.8.4...v1.9.1

1.8.4

What's Changed
* New Rule-Based Controllers by manjavacas in https://github.com/jajimer/sinergym/pull/204
* Sinergym 1.8.3 - year, month, day and hour in simulator output directly by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/200
* Sinergym 1.8.4 - Wrappers tests by AlejandroCN7 in https://github.com/jajimer/sinergym/pull/202


**Full Changelog**: https://github.com/jajimer/sinergym/compare/v1.8.2...v1.8.4

1.8.2

Patch Notes

Fixes 192
Fixes 191
Fixes 166
Fixes 194
Fixes 198
Fixes 155
Fixes 151

Changelog

- Added try_env.py to workspace in Dockerfile.
- Deleted default parameter for dictionary in Normalize wrapper and add that dictionary specification to load_model.py.
- Added conda environment file to configure automatically with python 3.9 version and Sinergym module (and pip dependencies).
- Fix datacenter-cool-X environments name in Sinergym register.
- Re-estructure documentation order, installation explanation re-estructured too, added config tests, more examples about how to use Sinergym.
- Added `year` field to Sinergym in order to define runperiods.
- Added runperiod extra_parameter (changes in common, controller, wrappers, rewards, eplus_env in order to adapt it to new year field in environmets).
- New documentation in extra configuration.
- README.md synchronized with the Sinergym documentation.
- Tests adapted to new version.
- Modify .devcontainer.
- Deleted default python version (python3 command) from Dockerfile.
- Now it is defined a base reward class, `BaseReward`, which receives the environment as input as well as other parameters to be used. Then, the `__call__` method calculates the reward, based on the attributes of the environment.

1.7.8

Patch Notes

Fixes 180
Fixes 185
Fixes 187
Fixes 182
Fixes 5
Fixes 189
Fixes 190

Changelog

- Added docs/build/ to .gitignore.
- Deleted docs/build/ from repository.
- Added gh-pages to commit and push documentation workflow.
- Edited documentation with this new feature.
- Edited the gh-pages branch by removing the entire repository except for the documentation compilation.
- Modified merge workflow, documentation build in folder called compilation in main branch and added new path to index.html.
- Update documentation with new functionality.
- Added a python script to load trained models in local storage.
- Added stable-baselines3 module to *test* tag in setup.py extra_requires packages.
- Changed names generated in `DRL_battery.py` in order to solve 187
- Added *act_repeat* and *max_data_store_num* parameters to env constructor (for simulator).
- Added checked new weathers using Sinergym automatic adaptation IDF (119)
- Added Eplus-datacenter-cool-X environments to available registered ID's.
- Added news Eplus-datacenter-cool-X ID's to documentation.
- Adapted tests in order to be independent between them.

1.7.7

Patch Notes

Fixes 169
Fixes 171
Fixes 173
Fixes 168
Fixes 175
Fixes 176
Fixes 178

- Updated RBC according to [FSEC](https://www.fsec.ucf.edu/en/publications/pdf/FSEC-CR-2010-13.pdf) standard setpoints. Now both outdoor air drybulb temperature and hour (day, evening, night) are considered in order to update setpoints.
- A new function has been added to utils/common.py in order to avoid repeated code when obtaining seasonal comfort ranges.
- This new function has been added in rewards.py methods as well as in the updated RBC in controllers.py.
- Changed name *.vscode* to *.vscode_conf* in order to not affect current Visual Studio Code IDE.
- Updated *merge_pr.yml* workflow in order to solve 168
- Deleted `ERROR: Energyplus Completed Succesfully` message in Energyplus log.
- Fixed duplicated `progress.csv` lines in Sinergym logger.
- Stable Baselines algorithms don't implement an `env.close()` when training ends. This `env.close()` has been added in `DRL_battery.py` accordingly.
- `DRL_battery.py` total timesteps training with MINUS 1 in order to not generate an empty episode at the end.
- Re-structuration of Evaluation Callback.
- Separated training env of evaluation env in `DRL_battery.py` in order to not merge logs and outputs of both processes.
- Fixed tests to integrate all bug fixes explained above.
- Modified MANIFEST.in in order to include *sinergym/data* in PyPi distribution and solve 176
- Solved 178

Page 7 of 9

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.