omicverse.pl.boxplot

Contents

omicverse.pl.boxplot#

omicverse.pl.boxplot(data, hue, x_value=None, y_value=None, width=0.3, title='', figsize=(6, 3), palette=None, fontsize=10, legend_bbox=(1, 0.55), legend_ncol=1, hue_order=None, *, x=None, y=None, ax=None, show_points=True)[source]#

Create a boxplot with jittered points to visualize data distribution across categories.

Parameters:

  • data (pd.DataFrame) – Input table containing grouping and numeric columns.

  • hue (str) – Column name used for color grouping.

  • x_value (str) – Column name used as x-axis category. x is accepted as an alias, which is what the sibling table-first plots (barplot, stripplot, violinplot) call it.

  • y_value (str) – Column name containing numeric values. y is accepted as an alias.

  • width (float) – Width of each box element.

  • title (str) – Plot title.

  • figsize (tuple) – Figure size passed to matplotlib. Ignored when ax is given — the axes already has a size, and resizing its figure would rescale every other panel sharing it.

  • palette (list or None) – Color list for hue groups; default palette is used when None.

  • fontsize (int) – Base font size for ticks/labels.

  • legend_bbox (tuple) – Legend anchor position.

  • legend_ncol (int) – Number of legend columns.

  • hue_order (list or None) – Explicit order of hue categories.

  • x (str) – Aliases for x_value / y_value. Keyword-only.

  • y (str) – Aliases for x_value / y_value. Keyword-only.

  • ax (matplotlib.axes.Axes or None) – Draw into this axes instead of creating a figure. Keyword-only, so every existing positional call is unaffected. Pass one of multipanel()’s panels here to place the boxplot inside a larger figure.

  • show_points (bool) – Overlay jittered raw points on each box. True (default) keeps the existing look; False draws boxes only, for a clean multi-panel figure.

Returns:

Figure and axes of generated boxplot. The pair is returned in both cases — when ax was supplied the figure is simply the one that axes already belongs to — so the return shape never depends on how the function was called.

Return type:

Tuple[matplotlib.figure.Figure, matplotlib.axes.Axes]

Contents