The list of priority improvements for the 3W Project that we intend to develop collaboratively with the community is detailed below.
- Extend the 3W Dataset with more instances of new event types;
- Finalize incorporation of MAIS into the 3W Toolkit;
- Evaluate and if appropriate start using Git LFS;
- Configure other GitHub resources that may be useful for our development. What resources exactly?
- Incorporate and provide in this repository documentation automatically generated from docstrings. How exactly?
- Review strategy for generating
folds_clf_XX.csv; - Review strategy for virtual environment specification (
environment.yml); - Develop a
setup.py. Is this module interesting for our project? - Develop tool to generate
diffbetween versions of the 3W Dataset - Improve presentation of the 3W Dataset citation list;
- Develop unit tests for the main methods and functions;
- Set up action for automatic execution of unit tests after creating PRs;
- Establish coding guidelines. Which one?
- Reevaluate the use of the rolling_window.py. Is there a better option or a newer version?
- Evaluate inclusion of specific features for hyperparameter optimization;
- Assess feasibility and benefits of using Sklearn Pipeline;
- Evaluate the use of Docker to facilitate the use of the 3W Toolkit and the approval of contributions;
- Establish one or more time-related metrics for anomaly detection and classification.