.. _immplscatterdensityrst:

===================
mpl-scatter-density
===================


.. only:: html

    **Links:** :download:`notebook <im_mpl_scatter_density.ipynb>`, :downloadlink:`html <im_mpl_scatter_density2html.html>`, :download:`PDF <im_mpl_scatter_density.pdf>`, :download:`python <im_mpl_scatter_density.py>`, :downloadlink:`slides <im_mpl_scatter_density.slides.html>`, :githublink:`GitHub|_doc/notebooks/2016/pydata/im_mpl_scatter_density.ipynb|*`


`mpl-scatter-density <https://github.com/astrofrog/mpl-scatter-density>`__
speeds up density graph. `matplotlib <https://matplotlib.org/>`__ is
very slow when it comes to draw millions of points.
`datashader <https://github.com/bokeh/datashader>`__ is one alternative
but was meant for zooming/dezooming. This package provides a simple
functionality. The example comes the documentation.

.. code:: 

    %matplotlib inline

.. code:: 

    from jyquickhelper import add_notebook_menu
    add_notebook_menu()






.. contents::
    :local:





example
-------

.. code:: 

    import numpy as np
    import mpl_scatter_density
    import matplotlib.pyplot as plt
    import matplotlib.colors as col
    
    # Generate fake data
    
    N = 10000000
    x = np.random.normal(4, 2, N)
    y = np.random.normal(3, 1, N)
    
    # Make the plot - note that for the projection option to work, the
    # mpl_scatter_density module has to be imported above.
    
    fig = plt.figure()
    ax = fig.add_subplot(1, 1, 1, projection='scatter_density')
    try:
        ax.scatter_density(x, y)
        ax.set_xlim(-5, 10)
        ax.set_ylim(-5, 10)
    except AttributeError as e:
        print('issue with more recent version')
        print(e)


.. parsed-literal::
    issue with more recent version
    'NoneType' object has no attribute 'vmin'



.. image:: im_mpl_scatter_density_4_1.png


The corresponding *matplotib* function does not exist as is. The module
`seaborn <https://seaborn.pydata.org/examples/multiple_joint_kde.html>`__
provides density visualization but it is not designed for such big
sample.

.. code:: 

    import numpy as np
    import seaborn as sns
    import matplotlib.pyplot as plt
    
    # Generate fake data
    
    N = 10000
    x = np.random.normal(4, 2, N)
    y = np.random.normal(3, 1, N)
    
    #import seaborn as sns
    import matplotlib.pyplot as plt
    f, ax = plt.subplots(figsize=(8, 8))
    
    ax = sns.kdeplot(x, y, shade=True, shade_lowest=False)


.. parsed-literal::
    c:\python372_x64\lib\site-packages\statsmodels\tools\_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.
      import pandas.util.testing as tm



.. image:: im_mpl_scatter_density_6_1.png