Release 0.2.0

[AnFreTh]

This pull request includes several changes to improve the installation process, update dependencies, and enhance the functionality of the stream_topic package. The most important changes include modifying the installation instructions, updating the setup.py file to handle additional requirements and post-installation tasks, and making various improvements to the models.

Installation and Dependencies:

• README.md: Updated the installation instructions to use stream-topic instead of stream_topic and added instructions for installing additional requirements for add-ons like visualizations, BERTopic, and DCTE.
• setup.py: Added a PostInstallCommand class to handle downloading NLTK resources after installation and included extras_require for optional dependencies. [1] [2] [3] [4]

Version Update:

• stream_topic/__version__.py: Updated the version from 0.1.9 to 0.2.0.

Model Improvements:

• stream_topic/metrics/coherence_metrics.py: Added a get_info method to provide information about the metric and removed the _load_default_texts function. [1] [2]
• stream_topic/models/DCTE.py: Implemented lazy imports for SetFit components to handle missing dependencies more gracefully and updated the fit method to use these components. [1] [2] [3]

Miscellaneous Changes:

• stream_topic/models/KmeansTM.py, stream_topic/models/bertopicTM.py, stream_topic/models/cbc.py: Updated the OneHotEncoder to use sparse_output=False instead of sparse=False for compatibility with newer versions of scikit-learn. [1] [2] [3]
• stream_topic/models/abstract_helper_models/base.py: Removed direct imports of optuna and added dynamic importing to handle missing dependencies. [1] [2]