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

.. only:: html

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

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

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

.. _sphx_glr_auto_tutorial_plot_gexternal_lightgbm_reg.py:


.. _example-lightgbm-reg:

Convert a pipeline with a LightGBM regressor
============================================

.. index:: LightGBM

The discrepancies observed when using float and TreeEnsemble operator
(see :ref:`l-example-discrepencies-float-double`)
explains why the converter for *LGBMRegressor* may introduce significant
discrepancies even when it is used with float tensors.

Library *lightgbm* is implemented with double. A random forest regressor
with multiple trees computes its prediction by adding the prediction of
every tree. After being converting into ONNX, this summation becomes
:math:`\left[\sum\right]_{i=1}^F float(T_i(x))`,
where *F* is the number of trees in the forest,
:math:`T_i(x)` the output of tree *i* and :math:`\left[\sum\right]`
a float addition. The discrepancy can be expressed as
:math:`D(x) = |\left[\sum\right]_{i=1}^F float(T_i(x)) -
\sum_{i=1}^F T_i(x)|`.
This grows with the number of trees in the forest.

To reduce the impact, an option was added to split the node
*TreeEnsembleRegressor* into multiple ones and to do a summation
with double this time. If we assume the node if split into *a* nodes,
the discrepancies then become
:math:`D'(x) = |\sum_{k=1}^a \left[\sum\right]_{i=1}^{F/a}
float(T_{ak + i}(x)) - \sum_{i=1}^F T_i(x)|`.

.. contents::
    :local:

Train a LGBMRegressor
+++++++++++++++++++++

.. GENERATED FROM PYTHON SOURCE LINES 41-64

.. code-block:: default

    import packaging.version as pv
    import warnings
    import timeit
    import numpy
    from pandas import DataFrame
    import matplotlib.pyplot as plt
    from tqdm import tqdm
    from lightgbm import LGBMRegressor
    from onnxruntime import InferenceSession
    from skl2onnx import to_onnx, update_registered_converter
    from skl2onnx.common.shape_calculator import calculate_linear_regressor_output_shapes  # noqa
    from onnxmltools import __version__ as oml_version
    from onnxmltools.convert.lightgbm.operator_converters.LightGbm import convert_lightgbm  # noqa


    N = 1000
    X = numpy.random.randn(N, 20)
    y = (numpy.random.randn(N) +
         numpy.random.randn(N) * 100 * numpy.random.randint(0, 1, 1000))

    reg = LGBMRegressor(n_estimators=1000)
    reg.fit(X, y)






.. raw:: html

    <div class="output_subarea output_html rendered_html output_result">
    <style>#sk-container-id-16 {color: black;background-color: white;}#sk-container-id-16 pre{padding: 0;}#sk-container-id-16 div.sk-toggleable {background-color: white;}#sk-container-id-16 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-16 label.sk-toggleable__label-arrow:before {content: "▸";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-16 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-16 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-16 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-16 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-16 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-16 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: "▾";}#sk-container-id-16 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-16 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-16 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-16 div.sk-parallel-item::after {content: "";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-16 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 div.sk-serial::before {content: "";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-16 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-16 div.sk-item {position: relative;z-index: 1;}#sk-container-id-16 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-16 div.sk-item::before, #sk-container-id-16 div.sk-parallel-item::before {content: "";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-16 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-16 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-16 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-16 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-16 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-16 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-16 div.sk-label-container {text-align: center;}#sk-container-id-16 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-16 div.sk-text-repr-fallback {display: none;}</style><div id="sk-container-id-16" class="sk-top-container"><div class="sk-text-repr-fallback"><pre>LGBMRegressor(n_estimators=1000)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class="sk-container" hidden><div class="sk-item"><div class="sk-estimator sk-toggleable"><input class="sk-toggleable__control sk-hidden--visually" id="sk-estimator-id-57" type="checkbox" checked><label for="sk-estimator-id-57" class="sk-toggleable__label sk-toggleable__label-arrow">LGBMRegressor</label><div class="sk-toggleable__content"><pre>LGBMRegressor(n_estimators=1000)</pre></div></div></div></div></div>
    </div>
    <br />
    <br />

.. GENERATED FROM PYTHON SOURCE LINES 65-76

Register the converter for LGBMClassifier
+++++++++++++++++++++++++++++++++++++++++

The converter is implemented in :epkg:`onnxmltools`:
`onnxmltools...LightGbm.py
<https://github.com/onnx/onnxmltools/blob/master/onnxmltools/convert/
lightgbm/operator_converters/LightGbm.py>`_.
and the shape calculator:
`onnxmltools...Regressor.py
<https://github.com/onnx/onnxmltools/blob/master/onnxmltools/convert/
lightgbm/shape_calculators/Regressor.py>`_.

.. GENERATED FROM PYTHON SOURCE LINES 76-97

.. code-block:: default



    def skl2onnx_convert_lightgbm(scope, operator, container):
        options = scope.get_options(operator.raw_operator)
        if 'split' in options:
            if pv.Version(oml_version) < pv.Version('1.9.2'):
                warnings.warn(
                    "Option split was released in version 1.9.2 but %s is "
                    "installed. It will be ignored." % oml_version)
            operator.split = options['split']
        else:
            operator.split = None
        convert_lightgbm(scope, operator, container)


    update_registered_converter(
        LGBMRegressor, 'LightGbmLGBMRegressor',
        calculate_linear_regressor_output_shapes,
        skl2onnx_convert_lightgbm,
        options={'split': None})








.. GENERATED FROM PYTHON SOURCE LINES 98-104

Convert
+++++++

We convert the same model following the two scenarios, one single
TreeEnsembleRegressor node, or more. *split* parameter is the number of
trees per node TreeEnsembleRegressor.

.. GENERATED FROM PYTHON SOURCE LINES 104-111

.. code-block:: default


    model_onnx = to_onnx(reg, X[:1].astype(numpy.float32),
                         target_opset={'': 14, 'ai.onnx.ml': 2})
    model_onnx_split = to_onnx(reg, X[:1].astype(numpy.float32),
                               target_opset={'': 14, 'ai.onnx.ml': 2},
                               options={'split': 100})








.. GENERATED FROM PYTHON SOURCE LINES 112-114

Discrepancies
+++++++++++++

.. GENERATED FROM PYTHON SOURCE LINES 114-130

.. code-block:: default


    sess = InferenceSession(model_onnx.SerializeToString())
    sess_split = InferenceSession(model_onnx_split.SerializeToString())

    X32 = X.astype(numpy.float32)
    expected = reg.predict(X32)
    got = sess.run(None, {'X': X32})[0].ravel()
    got_split = sess_split.run(None, {'X': X32})[0].ravel()

    disp = numpy.abs(got - expected).sum()
    disp_split = numpy.abs(got_split - expected).sum()

    print("sum of discrepancies 1 node", disp)
    print("sum of discrepancies split node",
          disp_split, "ratio:", disp / disp_split)





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

 .. code-block:: none

    sum of discrepancies 1 node 0.00012932610302943083
    sum of discrepancies split node 4.181837350948408e-05 ratio: 3.0925665485316083




.. GENERATED FROM PYTHON SOURCE LINES 131-133

The sum of the discrepancies were reduced 4, 5 times.
The maximum is much better too.

.. GENERATED FROM PYTHON SOURCE LINES 133-140

.. code-block:: default


    disc = numpy.abs(got - expected).max()
    disc_split = numpy.abs(got_split - expected).max()

    print("max discrepancies 1 node", disc)
    print("max discrepancies split node", disc_split, "ratio:", disc / disc_split)





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

 .. code-block:: none

    max discrepancies 1 node 1.307127771354999e-06
    max discrepancies split node 3.2222179946472806e-07 ratio: 4.056608750638187




.. GENERATED FROM PYTHON SOURCE LINES 141-145

Processing time
+++++++++++++++

The processing time is slower but not much.

.. GENERATED FROM PYTHON SOURCE LINES 145-153

.. code-block:: default


    print("processing time no split",
          timeit.timeit(
              lambda: sess.run(None, {'X': X32})[0], number=150))
    print("processing time split",
          timeit.timeit(
              lambda: sess_split.run(None, {'X': X32})[0], number=150))





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

 .. code-block:: none

    processing time no split 6.361445982940495
    processing time split 6.967872331850231




.. GENERATED FROM PYTHON SOURCE LINES 154-159

Split influence
+++++++++++++++

Let's see how the sum of the discrepancies moves against
the parameter *split*.

.. GENERATED FROM PYTHON SOURCE LINES 159-174

.. code-block:: default


    res = []
    for i in tqdm(list(range(20, 170, 20)) + [200, 300, 400, 500]):
        model_onnx_split = to_onnx(reg, X[:1].astype(numpy.float32),
                                   target_opset={'': 14, 'ai.onnx.ml': 2},
                                   options={'split': i})
        sess_split = InferenceSession(model_onnx_split.SerializeToString())
        got_split = sess_split.run(None, {'X': X32})[0].ravel()
        disc_split = numpy.abs(got_split - expected).max()
        res.append(dict(split=i, disc=disc_split))

    df = DataFrame(res).set_index('split')
    df["baseline"] = disc
    print(df)





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

 .. code-block:: none


      0%|          | 0/12 [00:00<?, ?it/s]
      8%|8         | 1/12 [00:16<03:01, 16.52s/it]
     17%|#6        | 2/12 [00:31<02:38, 15.87s/it]
     25%|##5       | 3/12 [00:47<02:22, 15.82s/it]
     33%|###3      | 4/12 [01:02<02:04, 15.55s/it]
     42%|####1     | 5/12 [01:18<01:49, 15.65s/it]
     50%|#####     | 6/12 [01:33<01:32, 15.47s/it]
     58%|#####8    | 7/12 [01:48<01:16, 15.36s/it]
     67%|######6   | 8/12 [02:04<01:01, 15.45s/it]
     75%|#######5  | 9/12 [02:19<00:45, 15.26s/it]
     83%|########3 | 10/12 [02:34<00:30, 15.33s/it]
     92%|#########1| 11/12 [02:49<00:15, 15.19s/it]
    100%|##########| 12/12 [03:04<00:00, 15.07s/it]
    100%|##########| 12/12 [03:04<00:00, 15.38s/it]
                   disc  baseline
    split                        
    20     1.771951e-07  0.000001
    40     2.930574e-07  0.000001
    60     2.208387e-07  0.000001
    80     3.101587e-07  0.000001
    100    3.222218e-07  0.000001
    120    3.905435e-07  0.000001
    140    3.101587e-07  0.000001
    160    3.184264e-07  0.000001
    200    4.363748e-07  0.000001
    300    5.597714e-07  0.000001
    400    7.869959e-07  0.000001
    500    7.729833e-07  0.000001




.. GENERATED FROM PYTHON SOURCE LINES 175-176

Graph.

.. GENERATED FROM PYTHON SOURCE LINES 176-182

.. code-block:: default

    _, ax = plt.subplots(1, 1)
    df.plot(title="Sum of discrepancies against split\n"
                  "split = number of tree per node",
            ax=ax)

    # plt.show()



.. image-sg:: /auto_tutorial/images/sphx_glr_plot_gexternal_lightgbm_reg_001.png
   :alt: Sum of discrepancies against split split = number of tree per node
   :srcset: /auto_tutorial/images/sphx_glr_plot_gexternal_lightgbm_reg_001.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** ( 8 minutes  19.778 seconds)


.. _sphx_glr_download_auto_tutorial_plot_gexternal_lightgbm_reg.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_gexternal_lightgbm_reg.py <plot_gexternal_lightgbm_reg.py>`

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

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


.. only:: html

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

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