.. only:: html

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

        Click :ref:`here <sphx_glr_download_auto_examples_plot_model_selection2.py>`     to download the full example code
    .. rst-class:: sphx-glr-example-title

    .. _sphx_glr_auto_examples_plot_model_selection2.py:


==========================
Hyperparameter Selection 2
==========================

This example demonstrates how to do model selection in a pipeline where segments are learned directly by a neural network



.. image:: /auto_examples/images/sphx_glr_plot_model_selection2_001.png
    :class: sphx-glr-single-img


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

 Out:

 .. code-block:: none

    /home/david/Code/venv3/lib/python3.6/site-packages/seglearn/split.py:100: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray
      Xt_new = np.array(Xt_new)
    Fitting 3 folds for each of 3 candidates, totalling 9 fits
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=100; total time=   2.4s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=100; total time=   2.6s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=100; total time=   2.7s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=200; total time=   2.8s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=200; total time=   2.6s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=200; total time=   2.8s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=400; total time=   3.1s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=400; total time=   3.2s
    [CV] END crnn__width=seg__width, seg__overlap=0.0, seg__width=400; total time=   3.1s
    /home/david/Code/seglearn/examples/plot_model_selection2.py:74: UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure.
      plt.show()






|


.. code-block:: default


    # Author: David Burns
    # License: BSD

    import matplotlib.pyplot as plt
    from tensorflow.python.keras.layers import Dense, LSTM, Conv1D
    from tensorflow.python.keras.models import Sequential
    from tensorflow.python.keras.wrappers.scikit_learn import KerasClassifier

    from sklearn.model_selection import GridSearchCV

    from seglearn.datasets import load_watch
    from seglearn.pipe import Pype
    from seglearn.split import TemporalKFold
    from seglearn.transform import Segment


    def crnn_model(width=100, n_vars=6, n_classes=7, conv_kernel_size=5,
                   conv_filters=2, lstm_units=2):
        # create a crnn model with keras with one cnn layers, and one rnn layer
        input_shape = (width, n_vars)
        model = Sequential()
        model.add(Conv1D(filters=conv_filters, kernel_size=conv_kernel_size,
                         padding='valid', activation='relu', input_shape=input_shape))
        model.add(LSTM(units=lstm_units, dropout=0.1, recurrent_dropout=0.1))
        model.add(Dense(n_classes, activation="softmax"))

        model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])

        return model


    # load the data
    data = load_watch()
    X = data['X']
    y = data['y']

    # temporal splitting of data
    splitter = TemporalKFold(n_splits=3)
    Xs, ys, cv = splitter.split(X, y)

    # create a segment learning pipeline
    pipe = Pype([('seg', Segment(order='C')),
                 ('crnn', KerasClassifier(build_fn=crnn_model, epochs=1, batch_size=256, verbose=0))])

    # create a parameter dictionary using the sklearn API
    #
    # you can also set a parameter to be always equal to another parameter, by setting its value to
    # parameter name to track (this is an extension from sklearn)
    #
    # note that if you want to set a parameter to a single value, it will still need to be as a list

    par_grid = {'seg__width': [100, 200, 400],
                'seg__overlap': [0.],
                'crnn__width': ['seg__width']}

    clf = GridSearchCV(pipe, par_grid, cv=cv, verbose=2)
    clf.fit(Xs, ys)
    scores = clf.cv_results_['mean_test_score']
    stds = clf.cv_results_['std_test_score']

    plt.plot(par_grid['seg__width'], scores, '-o')
    plt.title("Grid Search Scores")
    plt.xlabel("Width [s]")
    plt.ylabel("CV Average Score")
    plt.fill_between(par_grid['seg__width'], scores - stds, scores + stds, alpha=0.2, color='navy')
    plt.show()


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

   **Total running time of the script:** ( 0 minutes  27.853 seconds)


.. _sphx_glr_download_auto_examples_plot_model_selection2.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_model_selection2.py <plot_model_selection2.py>`



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

     :download:`Download Jupyter notebook: plot_model_selection2.ipynb <plot_model_selection2.ipynb>`


.. only:: html

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

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