feat: rebuild with pydantic

[kengz]

How does it work

Feature Transform simply builds sklearn ColumnTransformer and its estimators/pipelines with 1-1 mapping from a spec file:

1. Spec is defined via Pydantic feature_transform/validator/. This defines:
   • spec: the Estimator, Pipeline, ColumnTransformer
2. If spec specifies:
   1. transformers=list[(name, transformer, columns)], then use ColumnTransformer
   2. transformers=list[(transformer, columns)], then use make_column_transformer with auto-generated names

See more in the pydantic spec definition:

• feature_transform/validator/spec.py: the spec used by feature_transform

Guiding principles

The design of Feature Transform is guided as follows:

1. simple: the module spec is straightforward:
   1. it is simply sklearn class name with kwargs.
   2. it supports official sklearn estimators, Pipeline, and custom-defined modules registered via ft.register_class
2. expressive: it can be used to build both simple and advanced ColumnTransformer easily
3. portable: it returns ColumnTransformer that can be used anywhere; it is not a framework.
4. parametrizable: data-based feature transformation unlocks fast experimentation, e.g. by building logic for hyperparameter / data feature search

[github-actions]

🎉 This PR is included in version 1.0.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀