coco_pipe.fm.cbramod.pipeline
=============================

.. py:module:: coco_pipe.fm.cbramod.pipeline

.. autoapi-nested-parse::

   Pipeline wrapper for running regression with the CBRAMod foundation model.

   This mirrors the public surface of the classical regression pipeline but swaps
   the estimator for a foundation-model-backed regressor. The embedder is expected
   to turn raw inputs into embeddings; a light downstream regressor is trained on
   top of those embeddings.



Classes
-------

.. autoapisummary::

   coco_pipe.fm.cbramod.pipeline.FoundationRegressor
   coco_pipe.fm.cbramod.pipeline.CBRAModRegressionPipeline


Module Contents
---------------

.. py:class:: FoundationRegressor(embed_fn: Callable[[Any], numpy.ndarray], base_regressor: Optional[sklearn.base.BaseEstimator] = None, multioutput: bool = False)

   Bases: :py:obj:`sklearn.base.BaseEstimator`, :py:obj:`sklearn.base.RegressorMixin`


   Thin sklearn-compatible wrapper around a foundation model embedder.

   :param embed_fn: Callable that maps ``X`` to a 2D numpy array of embeddings. Either a
                    ``__call__`` or ``transform`` method will be used.
   :type embed_fn: callable
   :param base_regressor: Downstream regressor trained on the embeddings. Defaults to ``Ridge``.
   :type base_regressor: BaseEstimator, optional
   :param multioutput: If True, wraps the base regressor in ``MultiOutputRegressor`` for
                       multi-target regression.
   :type multioutput: bool, optional


   .. py:attribute:: embed_fn


   .. py:attribute:: base_regressor


   .. py:attribute:: multioutput
      :value: False



   .. py:method:: _embed(X: Any) -> numpy.ndarray


   .. py:method:: fit(X: Any, y: Any)


   .. py:method:: predict(X: Any) -> numpy.ndarray


.. py:class:: CBRAModRegressionPipeline(X: Union[pandas.DataFrame, numpy.ndarray], y: Union[pandas.Series, numpy.ndarray], embed_fn: Callable[[Any], numpy.ndarray], metrics: Union[str, Sequence[str], None] = None, base_regressor: Optional[sklearn.base.BaseEstimator] = None, hp_search_params: Optional[Dict[str, Sequence[Any]]] = None, use_scaler: bool = False, random_state: int = 42, n_jobs: int = -1, cv_kwargs: Optional[Dict[str, Any]] = None, groups: Optional[Union[pandas.Series, numpy.ndarray]] = None, verbose: bool = False)

   Bases: :py:obj:`coco_pipe.ml.base.BasePipeline`


   Regression pipeline that routes features through the CBRAMod foundation
   model before fitting a lightweight regressor.

   The interface mirrors ``RegressionPipeline`` for analysis_type dispatch but
   exposes a required ``embed_fn`` to obtain embeddings from the foundation
   model.


   .. py:attribute:: embed_fn


   .. py:attribute:: verbose
      :value: False



   .. py:attribute:: model_name
      :value: 'CBRAMod'



   .. py:method:: run(analysis_type: str = 'baseline', n_features: Optional[int] = None, direction: str = 'forward', search_type: str = 'grid', n_iter: int = 50, scoring: Optional[str] = None) -> Dict[str, Any]