.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/swissmetro/plot_b01logit_all_algos.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_swissmetro_plot_b01logit_all_algos.py>`
        to download the full example code

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

.. _sphx_glr_auto_examples_swissmetro_plot_b01logit_all_algos.py:


Logit model
===========

Estimation of a logit model with several algorithms.

:author: Michel Bierlaire, EPFL
:date: Tue Nov  7 17:00:09 2023

.. GENERATED FROM PYTHON SOURCE LINES 12-20

.. code-block:: default

    import itertools
    import pandas as pd
    from biogeme.tools import format_timedelta
    import biogeme.biogeme as bio
    from biogeme import models
    import biogeme.exceptions as excep
    from biogeme.expressions import Beta








.. GENERATED FROM PYTHON SOURCE LINES 21-22

See the data processing script: :ref:`swissmetro_data`.

.. GENERATED FROM PYTHON SOURCE LINES 22-36

.. code-block:: default

    from swissmetro_data import (
        database,
        CHOICE,
        SM_AV,
        CAR_AV_SP,
        TRAIN_AV_SP,
        TRAIN_TT_SCALED,
        TRAIN_COST_SCALED,
        SM_TT_SCALED,
        SM_COST_SCALED,
        CAR_TT_SCALED,
        CAR_CO_SCALED,
    )








.. GENERATED FROM PYTHON SOURCE LINES 37-38

Parameters to be estimated.

.. GENERATED FROM PYTHON SOURCE LINES 38-44

.. code-block:: default

    ASC_CAR = Beta('ASC_CAR', 0, None, None, 0)
    ASC_TRAIN = Beta('ASC_TRAIN', 0, None, None, 0)
    ASC_SM = Beta('ASC_SM', 0, None, None, 1)
    B_TIME = Beta('B_TIME', 0, None, None, 0)
    B_COST = Beta('B_COST', 0, None, None, 0)








.. GENERATED FROM PYTHON SOURCE LINES 45-46

Definition of the utility functions.

.. GENERATED FROM PYTHON SOURCE LINES 46-50

.. code-block:: default

    V1 = ASC_TRAIN + B_TIME * TRAIN_TT_SCALED + B_COST * TRAIN_COST_SCALED
    V2 = ASC_SM + B_TIME * SM_TT_SCALED + B_COST * SM_COST_SCALED
    V3 = ASC_CAR + B_TIME * CAR_TT_SCALED + B_COST * CAR_CO_SCALED








.. GENERATED FROM PYTHON SOURCE LINES 51-52

Associate utility functions with the numbering of alternatives.

.. GENERATED FROM PYTHON SOURCE LINES 52-54

.. code-block:: default

    V = {1: V1, 2: V2, 3: V3}








.. GENERATED FROM PYTHON SOURCE LINES 55-56

Associate the availability conditions with the alternatives.

.. GENERATED FROM PYTHON SOURCE LINES 56-58

.. code-block:: default

    av = {1: TRAIN_AV_SP, 2: SM_AV, 3: CAR_AV_SP}








.. GENERATED FROM PYTHON SOURCE LINES 59-61

Definition of the model. This is the contribution of each
observation to the log likelihood function.

.. GENERATED FROM PYTHON SOURCE LINES 61-63

.. code-block:: default

    logprob = models.loglogit(V, av, CHOICE)








.. GENERATED FROM PYTHON SOURCE LINES 64-66

Options for the optimization algorithm
--------------------------------------

.. GENERATED FROM PYTHON SOURCE LINES 68-69

The conjugate gradient iteration can be constrained to stay feasible, or not.

.. GENERATED FROM PYTHON SOURCE LINES 69-71

.. code-block:: default

    infeasible_cg_values = [True, False]








.. GENERATED FROM PYTHON SOURCE LINES 72-73

The radius of the first trust region is tested with three different values.

.. GENERATED FROM PYTHON SOURCE LINES 73-75

.. code-block:: default

    initial_radius_values = [0.1, 1.0, 10.0]








.. GENERATED FROM PYTHON SOURCE LINES 76-77

The percentage of iterations such that the analytical second derivatives is evaluated.

.. GENERATED FROM PYTHON SOURCE LINES 77-79

.. code-block:: default

    second_derivatives_values = [0.0, 0.5, 1.0]








.. GENERATED FROM PYTHON SOURCE LINES 80-81

We run the optimization algorithm with all possible combinations of the parameters. The results are stored in a Pandas DataFrame called ``summary``.

.. GENERATED FROM PYTHON SOURCE LINES 81-148

.. code-block:: default

    results = {}
    summary = pd.DataFrame(
        columns=[
            'LogLikelihood',
            'GradientNorm',
            'Optimization time',
            'TerminationCause',
            'Status',
        ]
    )

    for infeasible_cg, initial_radius, second_derivatives in itertools.product(
        infeasible_cg_values, initial_radius_values, second_derivatives_values
    ):
        # Create the Biogeme object
        the_biogeme = bio.BIOGEME(database, logprob, parameter_file='few_draws.toml')
        # We set the parameters of the optimization algorithm
        the_biogeme.infeasible_cg = infeasible_cg
        the_biogeme.initial_radius = initial_radius
        the_biogeme.second_derivatives = second_derivatives
        # We cancel the generation of the outputfiles
        the_biogeme.generate_html = False
        the_biogeme.generate_pickle = False

        name = (
            f'cg_{infeasible_cg}_radius_{initial_radius}_second_deriv_{second_derivatives}'
        )
        the_biogeme.modelName = f'b05normal_mixture_algo_{name}'.strip()
        result_data = {
            'InfeasibleCG': infeasible_cg,
            'InitialRadius': initial_radius,
            'SecondDerivatives': second_derivatives,
            'Status': 'Success',  # Assume success unless an exception is caught
        }

        try:
            results[name] = the_biogeme.estimate()
            opt_time = format_timedelta(
                results[name].data.optimizationMessages["Optimization time"]
            )

            result_data.update(
                {
                    'LogLikelihood': results[name].data.logLike,
                    'GradientNorm': results[name].data.gradientNorm,
                    'Optimization time': opt_time,
                    'TerminationCause': results[name].data.optimizationMessages[
                        "Cause of termination"
                    ],
                }
            )

        except excep.BiogemeError as e:
            print(e)
            result_data.update(
                {
                    'Status': 'Failed',
                    'LogLikelihood': None,
                    'GradientNorm': None,
                    'Optimization time': None,
                    'TerminationCause': str(e),
                }
            )
            results[name] = None

        summary = pd.concat([summary, pd.DataFrame([result_data])], ignore_index=True)





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

 .. code-block:: none

    /Users/bierlair/OnlineFiles/FilesOnGoogleDrive/github/biogeme/docs/examples/swissmetro/plot_b01logit_all_algos.py:146: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.
      summary = pd.concat([summary, pd.DataFrame([result_data])], ignore_index=True)




.. GENERATED FROM PYTHON SOURCE LINES 149-151

.. code-block:: default

    summary






.. raw:: html

    <div class="output_subarea output_html rendered_html output_result">
    <div>
    <style scoped>
        .dataframe tbody tr th:only-of-type {
            vertical-align: middle;
        }

        .dataframe tbody tr th {
            vertical-align: top;
        }

        .dataframe thead th {
            text-align: right;
        }
    </style>
    <table border="1" class="dataframe">
      <thead>
        <tr style="text-align: right;">
          <th></th>
          <th>LogLikelihood</th>
          <th>GradientNorm</th>
          <th>Optimization time</th>
          <th>TerminationCause</th>
          <th>Status</th>
          <th>InfeasibleCG</th>
          <th>InitialRadius</th>
          <th>SecondDerivatives</th>
        </tr>
      </thead>
      <tbody>
        <tr>
          <th>0</th>
          <td>-5331.252008</td>
          <td>0.024730</td>
          <td>184ms</td>
          <td>Relative gradient = 3.3e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>0.1</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>1</th>
          <td>-5331.252007</td>
          <td>0.017823</td>
          <td>44ms</td>
          <td>Relative gradient = 3e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>0.1</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>2</th>
          <td>-5331.252007</td>
          <td>0.000231</td>
          <td>55ms</td>
          <td>Relative gradient = 3.9e-08 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>0.1</td>
          <td>1.0</td>
        </tr>
        <tr>
          <th>3</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>134ms</td>
          <td>Relative gradient = 3.6e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>1.0</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>4</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>1.0</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>5</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>4ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>1.0</td>
          <td>1.0</td>
        </tr>
        <tr>
          <th>6</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>10.0</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>7</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>4ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>10.0</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>8</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>5ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>True</td>
          <td>10.0</td>
          <td>1.0</td>
        </tr>
        <tr>
          <th>9</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>0.1</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>10</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>0.1</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>11</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>0.1</td>
          <td>1.0</td>
        </tr>
        <tr>
          <th>12</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>1.0</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>13</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>4ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>1.0</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>14</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>1.0</td>
          <td>1.0</td>
        </tr>
        <tr>
          <th>15</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>10.0</td>
          <td>0.0</td>
        </tr>
        <tr>
          <th>16</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>10.0</td>
          <td>0.5</td>
        </tr>
        <tr>
          <th>17</th>
          <td>-5331.252008</td>
          <td>0.025605</td>
          <td>3ms</td>
          <td>Relative gradient = 4.8e-06 &lt;= 6.1e-06</td>
          <td>Success</td>
          <td>False</td>
          <td>10.0</td>
          <td>1.0</td>
        </tr>
      </tbody>
    </table>
    </div>
    </div>
    <br />
    <br />

.. GENERATED FROM PYTHON SOURCE LINES 152-155

.. code-block:: default

    SUMMARY_FILE = '01logit_all_algos.csv'
    summary.to_csv(SUMMARY_FILE, index=False)
    print(f'Summary reported in file {SUMMARY_FILE}')




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

 .. code-block:: none

    Summary reported in file 01logit_all_algos.csv





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

   **Total running time of the script:** (0 minutes 2.593 seconds)


.. _sphx_glr_download_auto_examples_swissmetro_plot_b01logit_all_algos.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_b01logit_all_algos.py <plot_b01logit_all_algos.py>`

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

      :download:`Download Jupyter notebook: plot_b01logit_all_algos.ipynb <plot_b01logit_all_algos.ipynb>`


.. only:: html

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

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