Function geoplot

missingno/missingno.py:583–807

Generates a geographical data nullity heatmap, which shows the distribution of missing data across geographic regions. The precise output depends on the inputs provided. In increasing order of usefulness: * If no geographical context is provided, a quadtree is computed and nullities

(df, x=None, y=None, coordinates=None, by=None, geometry=None, cutoff=None, histogram=False,
            figsize=(25, 10), fontsize=8, inline=True)

def geoplot(df, x=None, y=None, coordinates=None, by=None, geometry=None, cutoff=None, histogram=False,
            figsize=(25, 10), fontsize=8, inline=True):
    """
    Generates a geographical data nullity heatmap, which shows the distribution of missing data across geographic
    regions. The precise output depends on the inputs provided. In increasing order of usefulness:

    *   If no geographical context is provided, a quadtree is computed and nullities are rendered as abstract
        geopgrahical squares.
    *   If geographical context is provided in the form of a column of geographies (region, borough. ZIP code,
        etc.) in the `DataFrame`, convex hulls are computed for each of the point groups and the heatmap is generated
        within them.
    *   If geographical context is provided *and* a separate geometry is provided, a heatmap is generated for each
        point group within this geograpby instead.

    :param df: The DataFrame whose completeness is being mapped.
    :param x: The x variable: probably a coordinate (longitude), possibly some other floating point value. May be a
    string (pointing to a column of df) or an iterable.
    :param y: The y variable: probably a coordinate (latitude), possibly some other floating point value. May be a
    string (pointing to a column of df) or an iterable.
    :param coordinates: A coordinate tuple iterable, or column thereof in the given DataFrame. One of x AND y OR
    coordinates must be specified, but not both.
    :param by: If you would like to aggregate your geometry by some geospatial attribute of the underlying DataFrame,
    name that column here.
    :param geometry: If you would like to provide your own geometries for your aggregation, instead of relying on
    (functional, but not pretty) convex hulls, provide them here. This parameter is expected to be a dict or Series
    of `shapely.Polygon` or `shapely.MultiPolygon` objects. It's ignored if `by` is not specified.
    :param cutoff: If no aggregation is specified, this parameter sets the minimum number of observations to include in
    each square. If not provided, set to 50 or 5% of the total size of the dataset, whichever is smaller. If `by` is
    specified this parameter is ignored.
    :param figsize: The size of the figure to display. This is a `matplotlib` parameter which defaults to (25, 10).
    :param histogram: Whether or not to plot a histogram of data distributions below the map. Defaults to False.
    :param fontsize: If `hist` is specified, this parameter specifies the size of the tick labels. Ignored if `hist`
    is not specified. Defaults to 8.
    :param inline: Whether or not the figure is inline. If it's not then instead of getting plotted, this method will
    return its figure.
    :return: If `inline` is True, the underlying `matplotlib.figure` object. Else, nothing.
    """
    import shapely.geometry
    import descartes
    import matplotlib.cm
    # We produce a coordinate column in-place in a function-local copy of the `DataFrame`.
    # This seems specious, and sort of is, but is necessary because the internal `pandas` aggregation methods
    # (`pd.core.groupby.DataFrameGroupBy.count` specifically) are optimized to run two orders of magnitude faster than
    # user-defined external `groupby` operations. For example:
    # >>> %time df.head(100000).groupby(lambda ind: df.iloc[ind]['LOCATION']).count()
    # Wall time: 12.7 s
    # >>> %time df.head(100000).groupby('LOCATION').count()
    # Wall time: 96 ms
    x_col = '__x'
    y_col = '__y'
    if x and y:
        if isinstance(x, str) and isinstance(y, str):
            x_col = x
            y_col = y
        else:
            df['__x'] = x
            df['__y'] = y
    elif coordinates: