Function plotClusteredStackedBarchart

scripts/plotting/common.py:148–210

(dataToPlot,
                                 clusterLabels=None,
                                 clusterLabelPosition=(0.1, -0.05),
                                 categoryLabelPosition=(1.0, -0.05),
                                 title="multiple stacked bar plot",
                                 H="/",
                                 **kwargs)

# Taken from https://stackoverflow.com/questions/22787209/how-to-have-clusters-of-stacked-bars
def plotClusteredStackedBarchart(dataToPlot,
                                 clusterLabels=None,
                                 clusterLabelPosition=(0.1, -0.05),
                                 categoryLabelPosition=(1.0, -0.05),
                                 title="multiple stacked bar plot",
                                 H="/",
                                 **kwargs):
    n_df = len(dataToPlot)
    n_col = len(dataToPlot[0].columns)
    n_ind = len(dataToPlot[0].index)
    axe = plt.subplot(111)

    for df in dataToPlot:  # for each data frame
        axe = df.plot(kind="bar",
                      linewidth=0,
                      stacked=True,
                      ax=axe,
                      legend=False,
                      grid=False,
                      **kwargs)  # make bar plots

    subtractFromXOffset = 0
    if n_df <= 2:
        subtractFromXOffset = 0.10
    elif n_df <= 4:
        subtractFromXOffset = 0.15
    elif n_df <= 8:
        subtractFromXOffset = 0.2
    elif n_df <= 16:
        subtractFromXOffset = 0.22
        H = ''
    else:
        subtractFromXOffset = 0.23
        H = ''

    h, l = axe.get_legend_handles_labels()  # get the handles we want to modify
    for i in range(0, n_df * n_col, n_col):  # len(h) = n_col * n_df
        for j, pa in enumerate(h[i:i + n_col]):
            for rect in pa.patches:  # for each index
                rect.set_x(rect.get_x() + 1 / float(n_df + 1) * i / float(
                    n_col) - subtractFromXOffset)  # for 8 clusters subtract 0.15
                rect.set_hatch(2 * H * int(i / n_col))  # edited part
                rect.set_width(1 / float(n_df + 1))

    axe.set_xticks(-0.18 + (np.arange(0, 2 * n_ind, 2) + 1 / float(n_df + 1)) / 2.)
    axe.set_xticklabels(df.index, rotation=0)
    axe.set_title(title)

    # Add invisible data to add another legend
    n = []
    for i in range(n_df):
        n.append(axe.bar(0, 0, color="gray", hatch=H * i))

    # if categoryLabelPosition is not None:
    # l1 = axe.legend(h[:n_col], l[:n_col], ncol=1, bbox_to_anchor=categoryLabelPosition)
    # axe.add_artist(l1)

    if clusterLabels is not None and n_df <= 8 and clusterLabelPosition is not None: