.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "_auto_examples/kernel_ridge/plot_kernel_ridge_cv.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        Click :ref:`here <sphx_glr_download__auto_examples_kernel_ridge_plot_kernel_ridge_cv.py>`
        to download the full example code

.. rst-class:: sphx-glr-example-title

.. _sphx_glr__auto_examples_kernel_ridge_plot_kernel_ridge_cv.py:


Kernel ridge with cross-validation
==================================

This example demonstrates how to solve kernel ridge regression with a
cross-validation of the regularization parameter, using himalaya's estimator
``KernelRidgeCV``.

.. GENERATED FROM PYTHON SOURCE LINES 11-13

Create a random dataset
-----------------------

.. GENERATED FROM PYTHON SOURCE LINES 13-19

.. code-block:: default

    import numpy as np
    np.random.seed(0)
    n_samples, n_features, n_targets = 10, 20, 4
    X = np.random.randn(n_samples, n_features)
    Y = np.random.randn(n_samples, n_targets)








.. GENERATED FROM PYTHON SOURCE LINES 20-24

Limit of GridSearchCV
---------------------
In scikit-learn, one can use ``GridSearchCV`` to optimize hyperparameters
over cross-validation.

.. GENERATED FROM PYTHON SOURCE LINES 24-35

.. code-block:: default


    import sklearn.model_selection
    import sklearn.kernel_ridge

    estimator = sklearn.kernel_ridge.KernelRidge(kernel="linear")
    gscv = sklearn.model_selection.GridSearchCV(
        estimator=estimator,
        param_grid=dict(alpha=np.logspace(-2, 2, 5)),
    )
    gscv.fit(X, Y)





.. rst-class:: sphx-glr-script-out

 Out:

 .. code-block:: none


    GridSearchCV(estimator=KernelRidge(),
                 param_grid={'alpha': array([1.e-02, 1.e-01, 1.e+00, 1.e+01, 1.e+02])})



.. GENERATED FROM PYTHON SOURCE LINES 36-38

However, since ``GridSearchCV`` optimizes the average score over all targets,
it returns a single value for alpha.

.. GENERATED FROM PYTHON SOURCE LINES 38-40

.. code-block:: default

    gscv.best_params_





.. rst-class:: sphx-glr-script-out

 Out:

 .. code-block:: none


    {'alpha': 100.0}



.. GENERATED FROM PYTHON SOURCE LINES 41-45

KernelRidgeCV
-------------
To optimize each target independently, himalaya implements ``KernelRidgeCV``,
which supports any cross-validation scheme compatible with scikit-learn.

.. GENERATED FROM PYTHON SOURCE LINES 45-51

.. code-block:: default

    import himalaya.kernel_ridge

    model = himalaya.kernel_ridge.KernelRidgeCV(kernel="linear",
                                                alphas=np.logspace(-2, 2, 5))
    model.fit(X, Y)





.. rst-class:: sphx-glr-script-out

 Out:

 .. code-block:: none


    KernelRidgeCV(alphas=array([1.e-02, 1.e-01, 1.e+00, 1.e+01, 1.e+02]))



.. GENERATED FROM PYTHON SOURCE LINES 52-53

KernelRidgeCV returns a separate best alpha per target.

.. GENERATED FROM PYTHON SOURCE LINES 53-54

.. code-block:: default

    model.best_alphas_




.. rst-class:: sphx-glr-script-out

 Out:

 .. code-block:: none


    array([1.e+00, 1.e-02, 1.e+02, 1.e+02])




.. rst-class:: sphx-glr-timing

   **Total running time of the script:** ( 0 minutes  0.025 seconds)


.. _sphx_glr_download__auto_examples_kernel_ridge_plot_kernel_ridge_cv.py:


.. only :: html

 .. container:: sphx-glr-footer
    :class: sphx-glr-footer-example



  .. container:: sphx-glr-download sphx-glr-download-python

     :download:`Download Python source code: plot_kernel_ridge_cv.py <plot_kernel_ridge_cv.py>`



  .. container:: sphx-glr-download sphx-glr-download-jupyter

     :download:`Download Jupyter notebook: plot_kernel_ridge_cv.ipynb <plot_kernel_ridge_cv.ipynb>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_