.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/sampling/_plot_b03cnl.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_sampling__plot_b03cnl.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_sampling__plot_b03cnl.py:


Cross-nested logit
==================

Estimation of a cross-nested logit model using sampling of alternatives.

Michel Bierlaire
Sat Jul 26 2025, 14:57:17

.. GENERATED FROM PYTHON SOURCE LINES 11-42

.. code-block:: Python


    import pandas as pd
    from alternatives import (
        ID_COLUMN,
        all_alternatives,
        alternatives,
        asian_and_downtown,
        only_asian,
        only_downtown,
        partitions,
    )
    from compare import compare
    from IPython.core.display_functions import display
    from specification_sampling import V, combined_variables

    import biogeme.biogeme_logging as blog
    from biogeme.biogeme import BIOGEME
    from biogeme.expressions import Beta
    from biogeme.nests import NestsForCrossNestedLogit, OneNestForCrossNestedLogit
    from biogeme.results_processing import (
        EstimationResults,
        get_pandas_estimated_parameters,
    )
    from biogeme.sampling_of_alternatives import (
        ChoiceSetsGeneration,
        GenerateModel,
        SamplingContext,
        generate_segment_size,
    )
    from biogeme.tools import timeit


.. GENERATED FROM PYTHON SOURCE LINES 43-45

.. code-block:: Python

    logger = blog.get_screen_logger(level=blog.INFO)


.. GENERATED FROM PYTHON SOURCE LINES 46-54

.. code-block:: Python

    PARTITION = 'downtown'
    MEV_PARTITION = 'uniform_asian_or_downtown'
    SAMPLE_SIZE = 10  # out of 100 alternatives
    SAMPLE_SIZE_MEV = 63  # out of 63 alternatives
    CHOICE_COLUMN = 'cnl_3'
    MODEL_NAME = f'cnl_{SAMPLE_SIZE}_{SAMPLE_SIZE_MEV}'
    FILE_NAME = f'{MODEL_NAME}.dat'


.. GENERATED FROM PYTHON SOURCE LINES 55-59

.. code-block:: Python

    the_partition = partitions.get(PARTITION)
    if the_partition is None:
        raise ValueError(f'Unknown partition: {PARTITION}')


.. GENERATED FROM PYTHON SOURCE LINES 60-64

.. code-block:: Python

    segment_sizes = list(
        generate_segment_size(SAMPLE_SIZE, the_partition.number_of_segments())
    )


.. GENERATED FROM PYTHON SOURCE LINES 65-66

We use all alternatives in the nest.

.. GENERATED FROM PYTHON SOURCE LINES 66-73

.. code-block:: Python

    mev_partition = partitions.get(MEV_PARTITION)
    if mev_partition is None:
        raise ValueError(f'Unknown partition: {MEV_PARTITION}')
    mev_segment_sizes = [
        SAMPLE_SIZE_MEV,
    ]


.. GENERATED FROM PYTHON SOURCE LINES 74-75

Nests

.. GENERATED FROM PYTHON SOURCE LINES 77-78

Downtown

.. GENERATED FROM PYTHON SOURCE LINES 78-86

.. code-block:: Python

    mu_downtown = Beta('mu_downtown', 1, 1, None, 0)
    downtown_alpha_dict = {i: 0.5 for i in asian_and_downtown} | {
        i: 1 for i in only_downtown
    }
    downtown_nest = OneNestForCrossNestedLogit(
        nest_param=mu_downtown, dict_of_alpha=downtown_alpha_dict, name='downtown'
    )


.. GENERATED FROM PYTHON SOURCE LINES 87-88

Asian

.. GENERATED FROM PYTHON SOURCE LINES 88-99

.. code-block:: Python

    mu_asian = Beta('mu_asian', 1, 1, None, 0)
    asian_alpha_dict = {i: 0.5 for i in asian_and_downtown} | {i: 1.0 for i in only_asian}
    asian_nest = OneNestForCrossNestedLogit(
        nest_param=mu_asian, dict_of_alpha=asian_alpha_dict, name='asian'
    )

    cnl_nests = NestsForCrossNestedLogit(
        choice_set=all_alternatives,
        tuple_of_nests=(downtown_nest, asian_nest),
    )


.. GENERATED FROM PYTHON SOURCE LINES 100-102

.. code-block:: Python

    observations = pd.read_csv('obs_choice.dat')


.. GENERATED FROM PYTHON SOURCE LINES 103-118

.. code-block:: Python

    context = SamplingContext(
        the_partition=the_partition,
        sample_sizes=segment_sizes,
        individuals=observations,
        choice_column=CHOICE_COLUMN,
        alternatives=alternatives,
        id_column=ID_COLUMN,
        biogeme_file_name=FILE_NAME,
        utility_function=V,
        combined_variables=combined_variables,
        mev_partition=mev_partition,
        mev_sample_sizes=mev_segment_sizes,
        cnl_nests=cnl_nests,
    )


.. GENERATED FROM PYTHON SOURCE LINES 119-121

.. code-block:: Python

    logger.info(context.reporting())


.. GENERATED FROM PYTHON SOURCE LINES 122-125

.. code-block:: Python

    the_data_generation = ChoiceSetsGeneration(context=context)
    the_model_generation = GenerateModel(context=context)


.. GENERATED FROM PYTHON SOURCE LINES 126-128

.. code-block:: Python

    biogeme_database = the_data_generation.sample_and_merge(recycle=False)


.. GENERATED FROM PYTHON SOURCE LINES 129-131

.. code-block:: Python

    logprob = the_model_generation.get_cross_nested_logit()


.. GENERATED FROM PYTHON SOURCE LINES 132-135

.. code-block:: Python

    the_biogeme = BIOGEME(biogeme_database, logprob)
    the_biogeme.model_name = MODEL_NAME


.. GENERATED FROM PYTHON SOURCE LINES 136-137

Calculate the null log likelihood for reporting.

.. GENERATED FROM PYTHON SOURCE LINES 137-141

.. code-block:: Python

    the_biogeme.calculate_null_loglikelihood(
        {i: 1 for i in range(context.total_sample_size)}
    )


.. GENERATED FROM PYTHON SOURCE LINES 142-143

Estimate the parameters.

.. GENERATED FROM PYTHON SOURCE LINES 143-151

.. code-block:: Python

    try:
        results = EstimationResults.from_yaml_file(
            filename=f'saved_results/{the_biogeme.model_name}.yaml'
        )
    except FileNotFoundError:
        with timeit(f'Estimate of model {the_biogeme.model_name}'):
            results = the_biogeme.estimate()


.. GENERATED FROM PYTHON SOURCE LINES 152-154

.. code-block:: Python

    print(results.short_summary())


.. GENERATED FROM PYTHON SOURCE LINES 155-159

.. code-block:: Python

    parameters_tables = get_pandas_estimated_parameters(estimation_results=results)
    estimated_parameters = parameters_tables['Estimated parameters']
    display(estimated_parameters)


.. GENERATED FROM PYTHON SOURCE LINES 160-162

.. code-block:: Python

    df, msg = compare(estimated_parameters)


.. GENERATED FROM PYTHON SOURCE LINES 163-165

.. code-block:: Python

    print(df)


.. GENERATED FROM PYTHON SOURCE LINES 166-167

.. code-block:: Python

    print(msg)


.. _sphx_glr_download_auto_examples_sampling__plot_b03cnl.py:

.. only:: html

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

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

      :download:`Download Jupyter notebook: _plot_b03cnl.ipynb <_plot_b03cnl.ipynb>`

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

      :download:`Download Python source code: _plot_b03cnl.py <_plot_b03cnl.py>`

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

      :download:`Download zipped: _plot_b03cnl.zip <_plot_b03cnl.zip>`


.. only:: html

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

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