.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples\feature_selection\LassoFS.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_feature_selection_LassoFS.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_feature_selection_LassoFS.py:


Selective inference for Lasso Feature Selection
================================================
This example shows how to perform selective inference for Lasso feature selection using the `pythonsi` library. The method is based on the work by Duy et al. (2021)[1]. 
[1] Le Duy, V. N., & Takeuchi, I. (2021, March). Parametric programming approach for more powerful and general lasso selective inference. In International conference on artificial intelligence and statistics (pp. 901-909). PMLR.

.. GENERATED FROM PYTHON SOURCE LINES 7-18

.. code-block:: Python



    # Author: Tran Tuan Kiet

    from pythonsi import Pipeline
    from pythonsi.feature_selection import LassoFeatureSelection
    from pythonsi import Data
    from pythonsi.test_statistics import FSTestStatistic
    import numpy as np
    import matplotlib.pyplot as plt








.. GENERATED FROM PYTHON SOURCE LINES 19-21

Define the pipeline
-------------------

.. GENERATED FROM PYTHON SOURCE LINES 21-36

.. code-block:: Python



    def LassoFS() -> Pipeline:
        x = Data()
        y = Data()

        lasso = LassoFeatureSelection(lambda_=10)
        active_set = lasso.run(x, y)
        return Pipeline(
            inputs=(x, y), output=active_set, test_statistic=FSTestStatistic(x=x, y=y)
        )


    my_pipeline = LassoFS()








.. GENERATED FROM PYTHON SOURCE LINES 37-39

Generate data
--------------

.. GENERATED FROM PYTHON SOURCE LINES 39-53

.. code-block:: Python



    def gen_data(n, p, true_beta):
        x = np.random.normal(loc=0, scale=1, size=(n, p))
        true_beta = true_beta.reshape(-1, 1)

        mu = x.dot(true_beta)
        Sigma = np.identity(n)
        Y = mu + np.random.normal(loc=0, scale=1, size=(n, 1))
        return x, Y, Sigma


    x, y, sigma = gen_data(150, 5, np.asarray([0, 0, 0, 0, 0]))








.. GENERATED FROM PYTHON SOURCE LINES 54-56

Run the pipeline
-----------------

.. GENERATED FROM PYTHON SOURCE LINES 56-61

.. code-block:: Python


    selected_features, p_values = my_pipeline([x, y], sigma)
    print("Selected features: ", selected_features)
    print("P-values: ", p_values)





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

 .. code-block:: none

    Selected features:  [2 3 4]
    P-values:  [0.15210666782406781, 0.7928001593659839, 0.5567564061709589]




.. GENERATED FROM PYTHON SOURCE LINES 62-63

Plot the p-values

.. GENERATED FROM PYTHON SOURCE LINES 63-68

.. code-block:: Python

    plt.figure()
    plt.bar(range(len(p_values)), p_values)
    plt.xlabel("Feature index")
    plt.ylabel("P-value")
    plt.show()



.. image-sg:: /auto_examples/feature_selection/images/sphx_glr_LassoFS_001.png
   :alt: LassoFS
   :srcset: /auto_examples/feature_selection/images/sphx_glr_LassoFS_001.png
   :class: sphx-glr-single-img






.. _sphx_glr_download_auto_examples_feature_selection_LassoFS.py:

.. only:: html

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

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

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

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

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

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: LassoFS.zip <LassoFS.zip>`


.. only:: html

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

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