stub

Create margin annotations for data outside axes viewport.

Functions

`rescale_clip_outliers`([ax, ...])	Rescale the axes of a Matplotlib axes with plotted points to exclude outliers.
`stub_all_clipped_values`(ax, *[, marker_kws, ...])	Draw edge markers for points lying outside the plot's bounds.
`stub_edge_mark`(ax, x, y, *[, marker_kws, offset])	Draw an edge marker for a single clipped point.

Classes

CalcBoundsIQR

Functor to set outlier bounds a fixed ratio above/below interquartile range.

class CalcBoundsIQR[source]

Functor to set outlier bounds a fixed ratio above/below interquartile range.

Bounds are calculated as (q1 - iqr_multiplier * iqr, q3 + iqr_multiplier * iqr), where q1 and q3 are the first and third quartiles, respectively, and iqr is the interquartile range (q3 - q1).

__init__(iqr_multiplier: float = 1.5) → None[source]

Initialize functor.

Parameters:: iqr_multiplier (float, default 1.5) – The multiplier applied to the IQR to determine the bounds.

__call__(data: Sequence[float]) → Tuple[float, float][source]

Calculate the lower and upper bounds of the input data using the IQR method.

Parameters:: data (np.ndarray) – The input data for which the bounds are calculated.
Returns:: bounds – Lower and upper calculated bounds.
Return type:: Tuple[float, float]

rescale_clip_outliers(ax: Axes | None = None, calc_outlier_bounds: Callable | Tuple[Callable, Callable] | None = None, *, pad: float = 0.1) → None[source]

Rescale the axes of a Matplotlib axes with plotted points to exclude outliers.

This function inspects plotted points already on the axes object.

Parameters:

ax (mpl_Axes, optional) –
The Matplotlib Axes object to rescale.

If None, plt.gca() will be used.
calc_outlier_bounds (Union[Callable, Tuple[Callable, Callable]], optional) –
A function or a tuple of two functions to calculate the outlier bounds for the x and y data points.

Each function should accept an array of data points and return a tuple of (lower, upper) bounds. If None, defaults to interquartile range-based calculation (CalcBoundsIQR with multiplier 1.5).
pad (float, default 0.1) – How far should axes limits be padded beyond the non-outlier data range?

Return type:

None

Notes

In cases where there are no outliers or all points are outliers, axes limits will not be adjusted In cases where non-outliers take on only a single value, padding will be applied relative to that value or, if zero, one.

stub_all_clipped_values(ax: Axes, *, marker_kws: Dict = frozendict.frozendict({}), offset: float = 0.1) → None[source]

Draw edge markers for points lying outside the plot’s bounds.

The original point and associated error bar (if any) is moved into the margin with a note indicating the original location. If the point is out of xlim and out of ylim, the point will be retrieved with two edge markers (horizontal and vertical).

Parameters:

ax (matplotlib.axes.Axes) – The axes object to draw markers on.
marker_kws (dict, optional) – Keyword arguments forwarded to matplotlib plot for edge markers.
offset (float, default 0.1) – How far outside axis viewport to draw edge markers, proportional to axis viewport height or width.

Notes

Full error bar support for values below lower {x,y}lim remains be implemented. Current implementation assumes each error bar is associated with a scatter point.

stub_edge_mark(ax: Axes, x: float, y: float, *, marker_kws: Dict = frozendict.frozendict({}), offset: float = 0.1) → Tuple[float, float][source]

Draw an edge marker for a single clipped point.

Parameters:

ax (matplotlib.axes.Axes) – The axes object to draw markers on.
x (float) – The x and y coordinates of the point.
y (float) – The x and y coordinates of the point.
marker_kws (dict, optional) – Keyword arguments forwarded to matplotlib plot for edge markers.
offset (float, default 0.1) – How far outside axis viewport to draw edge markers, proportional to axis viewport height or width.

Returns:

edge_x, edge_y – The x and y coordinates of the placed edge marker.

Return type:

float