.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/plot_cran.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_plot_cran.py>`
        to download the full example code or to run this example in your browser via JupyterLite.

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

.. _sphx_glr_auto_examples_plot_cran.py:


Loading a RDA file with custom types from CRAN
==============================================

A more advanced example showing how to read a dataset in the RDATA format from
the CRAN repository of R packages that include custom R types.

.. GENERATED FROM PYTHON SOURCE LINES 11-30

We will show how to load the graph of the classical
`seven bridges of Königsberg problem
<https://en.wikipedia.org/wiki/Seven_Bridges_of_K%C3%B6nigsberg>`_ from the
R package igraphdata.

.. warning::
    This is for illustration purposes only. If you plan to use the same
    dataset repeatedly it is better to download it, or to use a package that
    caches it, such as
    `scikit-datasets <https://daviddiazvico.github.io/scikit-datasets/>`_.

We will make use of the function
:external+python:func:`urllib.request.urlopen` to load the url, as well as
the package rdata.
The package is a tar file so we need also to import the
:external+python:mod:`tarfile` module.
We will use the package `igraph <https://python.igraph.org/en/stable/>`_ for
constructing the graph in Python.
Finally, we will import some plotting routines from Matplotlib.

.. GENERATED FROM PYTHON SOURCE LINES 30-41

.. code-block:: Python


    import tarfile
    from urllib.request import urlopen

    import igraph
    import igraph.drawing
    import matplotlib.pyplot as plt
    from matplotlib.colors import to_hex

    import rdata








.. GENERATED FROM PYTHON SOURCE LINES 42-43

The following URL contains the link to download the package from CRAN.

.. GENERATED FROM PYTHON SOURCE LINES 43-49

.. code-block:: Python


    pkg_url = (
        "https://cran.r-project.org/src/contrib/Archive/"
        "igraphdata/igraphdata_1.0.0.tar.gz"
    )








.. GENERATED FROM PYTHON SOURCE LINES 50-54

The dataset is contained in the "data" folder, as it is common for
R packages.
The file is named Koenisberg and it is in the RDATA format
(.rda extension).

.. GENERATED FROM PYTHON SOURCE LINES 54-56

.. code-block:: Python

    data_path = "igraphdata/data/Koenigsberg.rda"








.. GENERATED FROM PYTHON SOURCE LINES 57-60

We proceed to open the package using
:external+python:func:`~urllib.request.urlopen`
and :external+python:mod:`tarfile`.

.. GENERATED FROM PYTHON SOURCE LINES 60-72

.. code-block:: Python



    with urlopen(pkg_url) as package:
        with tarfile.open(fileobj=package, mode="r|gz") as package_tar:
            for member in package_tar:
                if member.name == data_path:
                    dataset = package_tar.extractfile(member)
                    assert dataset
                    with dataset:
                        parsed = rdata.parser.parse_file(dataset)
                    break








.. GENERATED FROM PYTHON SOURCE LINES 73-74

We could try to convert this dataset to Python objects.

.. GENERATED FROM PYTHON SOURCE LINES 74-78

.. code-block:: Python


    converted = rdata.conversion.convert(parsed)
    print(converted)





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

 .. code-block:: none

    /home/docs/checkouts/readthedocs.org/user_builds/rdata/envs/latest/lib/python3.11/site-packages/rdata/conversion/_conversion.py:900: UserWarning: Missing constructor for R class "igraph". The underlying R object is returned instead.
      warnings.warn(
    {'Koenigsberg': [array([4.]), array([False]), array([1., 1., 3., 3., 3., 2., 2.]), array([0., 0., 0., 1., 2., 1., 1.]), array([1., 0., 6., 5., 2., 3., 4.]), array([1., 0., 2., 6., 5., 3., 4.]), array([0., 0., 2., 4., 7.]), array([0., 3., 6., 7., 7.]), [array([1., 0., 1.]), {np.str_('name'): array(['The seven bidges of Koenigsberg'], dtype='<U31')}, {np.str_('name'): array(['Altstadt-Loebenicht', 'Kneiphof', 'Vorstadt-Haberberg', 'Lomse'],
          dtype='<U19'), np.str_('Euler_letter'): array(['B', 'A', 'C', 'D'], dtype='<U1')}, {np.str_('Euler_letter'): array(['a', 'b', 'f', 'e', 'g', 'c', 'd'], dtype='<U1'), np.str_('name'): array(['Kraemer Bruecke', 'Schmiedebruecke', 'Holzbruecke',
           'Honigbruecke', 'Hohe Bruecke', 'Gruene Bruecke', 'Koettelbruecke'],
          dtype='<U15')}], REnvironment({}, REnvironment({}))]}




.. GENERATED FROM PYTHON SOURCE LINES 79-83

From this representation, we can see that .rda files contain a mapping
of variable names to objects, and not just one object as .rds files.
In this case there is just one variable called "Koenigsberg", as the
dataset itself, but that is not necessarily always the case.

.. GENERATED FROM PYTHON SOURCE LINES 85-89

We can also see that there is no default conversion for the "igraph"
class, representing a graph.
Thus, the converted object is a list of the underlying vectors used
by this type.

.. GENERATED FROM PYTHON SOURCE LINES 91-96

It is however possible to define our own conversion routines for R classes
using the package rdata.
For that purpose we need to create a "constructor" function, that accepts
as arguments the underlying object to convert and its attributes, and
returns the converted object.

.. GENERATED FROM PYTHON SOURCE LINES 98-107

In this example, the object will be received as a list, corresponding to
the `igraph_t structure defined by the igraph package
<https://github.com/igraph/igraph/blob/
50d46370fd677128cf758e4dd5c1de61dae9a3ef/include/
igraph_datatype.h#L110-L121>`_.
We will convert it to a :external+igraph:class:`~igraph.Graph` object from
the `Python version of the igraph package
<https://python.igraph.org/en/stable/>`_.
The attrs dict is empty and will not be used.

.. GENERATED FROM PYTHON SOURCE LINES 107-135

.. code-block:: Python



    def graph_constructor(obj, attrs):
        """Construct graph object from R representation."""
        n_vertices = int(obj[0][0])
        is_directed = obj[1]
        edge_from = obj[2].astype(int)
        edge_to = obj[3].astype(int)

        # output_edge_index = obj[4]
        # input_edge_index = obj[5]
        # output_vertex_edge_index = obj[6]
        # input_vertex_edge_index = obj[7]

        graph_attrs = obj[8][1]
        vertex_attrs = obj[8][2]
        edge_attrs = obj[8][3]

        return igraph.Graph(
            n=n_vertices,
            directed=is_directed,
            edges=list(zip(edge_from, edge_to, strict=True)),
            graph_attrs=graph_attrs,
            vertex_attrs=vertex_attrs,
            edge_attrs=edge_attrs,
        )









.. GENERATED FROM PYTHON SOURCE LINES 136-142

We create a dict with all the constructors that we want to apply.
In this case, we include first the default constructors (which
provide transformations for common R classes) and our newly created
constructor.
The key used for the dictionary entries should be the name of the
corresponding R class.

.. GENERATED FROM PYTHON SOURCE LINES 142-147

.. code-block:: Python

    constructor_dict = {
        **rdata.conversion.DEFAULT_CLASS_MAP,
        "igraph": graph_constructor,
    }








.. GENERATED FROM PYTHON SOURCE LINES 148-150

We can now call the :func:`rdata.conversion.convert` functtion, supplying
the dictionary of constructors to use.

.. GENERATED FROM PYTHON SOURCE LINES 150-152

.. code-block:: Python

    converted = rdata.conversion.convert(parsed, constructor_dict=constructor_dict)








.. GENERATED FROM PYTHON SOURCE LINES 153-155

Finally, we check the constructed graph by plotting it using the
:external+igraph:func:`igraph.drawing.plot` function.

.. GENERATED FROM PYTHON SOURCE LINES 155-168

.. code-block:: Python

    fig, axes = plt.subplots()
    plt.subplots_adjust(left=0, right=1, bottom=0, top=1)
    igraph.drawing.plot(
        converted["Koenigsberg"],
        target=axes,
        vertex_label=converted["Koenigsberg"].vs["name"],
        vertex_label_size=8,
        vertex_size=120,
        vertex_color=to_hex("tab:blue"),
        edge_label=converted["Koenigsberg"].es["name"],
        edge_label_size=8,
    )
    plt.show()



.. image-sg:: /auto_examples/images/sphx_glr_plot_cran_001.png
   :alt: plot cran
   :srcset: /auto_examples/images/sphx_glr_plot_cran_001.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** (0 minutes 1.229 seconds)


.. _sphx_glr_download_auto_examples_plot_cran.py:

.. only:: html

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

    .. container:: lite-badge

      .. image:: images/jupyterlite_badge_logo.svg
        :target: ../lite/lab/index.html?path=auto_examples/plot_cran.ipynb
        :alt: Launch JupyterLite
        :width: 150 px

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

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

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

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

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

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


.. only:: html

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

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