Function bar

missingno/missingno.py:307–378

Plots a bar chart of data nullities by column. :param df: The DataFrame whose completeness is being nullity matrix mapped. :param log: Whether or not to display a logorithmic plot. Defaults to False (linear). :param filter: The filter to apply to the heatmap. Should be one of "top"

(df, figsize=(24, 10), fontsize=16, labels=None, log=False, color=(0.25, 0.25, 0.25), inline=True,
        filter=None, n=0, p=0, sort=None)

def bar(df, figsize=(24, 10), fontsize=16, labels=None, log=False, color=(0.25, 0.25, 0.25), inline=True,
        filter=None, n=0, p=0, sort=None):
    """
    Plots a bar chart of data nullities by column.

    :param df: The DataFrame whose completeness is being nullity matrix mapped.
    :param log: Whether or not to display a logorithmic plot. Defaults to False (linear).
    :param filter: The filter to apply to the heatmap. Should be one of "top", "bottom", or None (default). See
    `nullity_filter()` for more information.
    :param n: The cap on the number of columns to include in the filtered DataFrame. See  `nullity_filter()` for
    more information.
    :param p: The cap on the percentage fill of the columns in the filtered DataFrame. See  `nullity_filter()` for
    more information.
    :param sort: The sort to apply to the heatmap. Should be one of "ascending", "descending", or None. See
    `nullity_sort()` for more information.
    :param figsize: The size of the figure to display. This is a `matplotlib` parameter. Defaults to (24,
    10).
    :param fontsize: The figure's font size. This default to 16.
    :param labels: Whether or not to display the column names. Would need to be turned off on particularly large
    displays. Defaults to True.
    :param color: The color of the filled columns. Default is a medium dark gray: the RGB multiple `(0.25, 0.25, 0.25)`.
    :return: If `inline` is True, the underlying `matplotlib.figure` object. Else, nothing.
    """
    # Get counts.
    nullity_counts = len(df) - df.isnull().sum()

    # Apply filters and sorts.
    df = nullity_filter(df, filter=filter, n=n, p=p)
    df = nullity_sort(df, sort=sort)

    # Create the basic plot.
    fig = plt.figure(figsize=figsize)
    (nullity_counts / len(df)).plot(kind='bar', figsize=figsize, fontsize=fontsize, color=color, log=log)

    # Get current axis.
    ax1 = plt.gca()

    # Start appending elements, starting with a modified bottom x axis.
    if labels or (labels is None and len(df.columns) <= 50):
        pos = ax1.get_xticks()
        ax1.set_xticklabels(ax1.get_xticklabels(), rotation=45, ha='right', fontsize=fontsize)

        # Create the numerical ticks.
        ax2 = ax1.twinx()
        if not log:
            # Simple if the plot is ordinary.
            ax2.set_yticks(ax1.get_yticks())
            ax2.set_yticklabels([int(n*len(df)) for n in ax1.get_yticks()], fontsize=fontsize)
        else:
            # For some reason when a logarithmic plot is specified `ax1` always contains two more ticks than actually
            # appears in the plot. For example, if we do `msno.histogram(collisions.sample(500), log=True)` the contents
            # of the naive `ax1.get_yticks()` is [1.00000000e-03, 1.00000000e-02, 1.00000000e-01, 1.00000000e+00,
            # 1.00000000e+01]. The fix is to ignore the first and last entries.
            #
            # Also note that when a log scale is used, we have to make it match the `ax1` layout ourselves.
            ax2.set_yscale('log')
            ax2.set_ylim(ax1.get_ylim())
            ax2.set_yticks(ax1.get_yticks()[1:-1])