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

.. only:: html

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

        :ref:`Go to the end <sphx_glr_download_auto_examples_plot_regressor.py>`
        to download the full example code

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

.. _sphx_glr_auto_examples_plot_regressor.py:


=============================================
Sparse Group Lasso for grouped sparse signals
=============================================

Estimates a Sparse Group Lasso model on a simulated sparse signal with high
group-level sparsity. The prediction is compared to the ground truth and to
the results of a Lasso model.

.. GENERATED FROM PYTHON SOURCE LINES 11-90



.. image-sg:: /auto_examples/images/sphx_glr_plot_regressor_001.png
   :alt: plot regressor
   :srcset: /auto_examples/images/sphx_glr_plot_regressor_001.png
   :class: sphx-glr-single-img


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

 .. code-block:: none

    Lasso performance:
    train: R^2 = 0.491, RMSE = 503.387
    test:  R^2 = 0.160, RMSE = 623.359

    SGL performance:
    train: R^2 = 0.575, RMSE = 460.063
    test:  R^2 = 0.352, RMSE = 547.717

    <matplotlib.legend.Legend object at 0x7fee34689100>





|

.. code-block:: default

    import groupyr as gpr
    import matplotlib.pyplot as plt
    import numpy as np

    from sklearn.model_selection import KFold
    from sklearn.linear_model import LassoCV
    from sklearn.metrics import r2_score, mean_squared_error
    from sklearn.model_selection import train_test_split


    X, y, groups, coef = gpr.datasets.make_group_regression(
        n_samples=400,
        n_groups=50,
        n_informative_groups=5,
        n_features_per_group=20,
        n_informative_per_group=18,
        noise=500,
        coef=True,
        random_state=10,
    )

    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=0.25, random_state=10
    )

    # Fit an SGL model for this data
    cv = KFold(random_state=1729, shuffle=True)
    sgl = gpr.SGLCV(
        groups=groups, cv=cv, l1_ratio=[0.0, 1.0], tuning_strategy="bayes", n_bayes_iter=50
    ).fit(X_train, y_train)

    # Fit a Lasso model on this data
    cv = KFold(random_state=1729, shuffle=True)
    lasso = LassoCV(cv=cv).fit(X_train, y_train)

    # Print model performance
    print("Lasso performance:")
    print(
        "train: R^2 = {0:5.3f}, RMSE = {1:7.3f}".format(
            r2_score(y_train, lasso.predict(X_train)),
            np.sqrt(mean_squared_error(y_train, lasso.predict(X_train))),
        )
    )
    print(
        "test:  R^2 = {0:5.3f}, RMSE = {1:7.3f}".format(
            r2_score(y_test, lasso.predict(X_test)),
            np.sqrt(mean_squared_error(y_test, lasso.predict(X_test))),
        )
    )
    print("\nSGL performance:")
    print(
        "train: R^2 = {0:5.3f}, RMSE = {1:7.3f}".format(
            r2_score(y_train, sgl.predict(X_train)),
            np.sqrt(mean_squared_error(y_train, sgl.predict(X_train))),
        )
    )
    print(
        "test:  R^2 = {0:5.3f}, RMSE = {1:7.3f}".format(
            r2_score(y_test, sgl.predict(X_test)),
            np.sqrt(mean_squared_error(y_test, sgl.predict(X_test))),
        )
    )

    # Plot predicted values
    plt.plot(y_test, sgl.predict(X_test), marker="o", ls="", alpha=0.7, label="SGL")
    plt.plot(y_test, lasso.predict(X_test), marker="o", ls="", alpha=0.7, label="LASSO")

    max_val = np.max(
        [np.max(y_test), np.max(lasso.predict(X_test)), np.max(sgl.predict(X_test))]
    )
    min_val = np.min(
        [np.min(y_test), np.min(lasso.predict(X_test)), np.min(sgl.predict(X_test))]
    )

    plt.plot([min_val, max_val], [min_val, max_val], ls=":", lw=2, color="black")
    plt.xlabel("Target values")
    plt.ylabel("Predicted values")

    plt.legend()


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

   **Total running time of the script:** ( 1 minutes  11.729 seconds)


.. _sphx_glr_download_auto_examples_plot_regressor.py:

.. only:: html

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




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

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

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

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


.. only:: html

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

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