soft_boundary builtin

soft_boundary(
    coords,
    values,
    *,
    domains=None,
    domain_column=None,
    target,
    buffer,
    weights=None,
    quantiles=Ellipsis,
)

Hard (target domain alone) vs soft statistics: also folding in samples of other domains within buffer, as a search opened with Search(..., soft=buffer) would.

Parameters:

Name	Type	Description	Default
coords	PointSet or array_like	Samples, or their (n, 2) or (n, 3) coordinates.	required
values	array_like or str	Values, or their column in coords; NaN is skipped.	required
domains	array_like	Domain of each sample, int or str.	None
domain_column	str	Column of coords holding the domains; give it or domains.	None
target	int or str	The domain whose statistics are computed.	required
buffer	float	Distance within which a sample of another domain is folded into the soft row.	required
weights	array_like or str	Declustering weights; default 1.	None
quantiles	sequence of float	Probabilities of the quantile columns.	Ellipsis

Returns:

Name	Type	Description
added	ndarray of bool	One per sample of the other domains, in their original order: whether it is within buffer of target and folded into the soft row.
table	Table	kind ("hard", "soft"), n, mean, variance, std, cv, min, max and one column per quantile named P10, P97.5, ...

used in

• soft-boundary statistics