data_spacing builtin

data_spacing(
    targets,
    data,
    search,
    *,
    composite_length=None,
    holes=None,
    n=None,
    hull=False,
)

Equivalent data spacing (Cabral Pinto & Deutsch, 2017).

With a search, sqrt(V / (c n)): V the volume of the search ellipsoid, n the composites inside it around the target and c the composite length. Without, the plan form: the mean over n of sqrt(pi r^2 / n), r the mean plan distance to the n-th and n + 1-th nearest samples. Holes on a square grid of spacing s give about s in both forms.

Parameters:

Name	Type	Description	Default
targets	(array_like, PointSet or BlockModel)	Locations to measure at, e.g. block centroids.	required
data	(Drillholes, PointSet or array_like)	Composites, or their (n, 2) or (n, 3) coordinates; a Drillholes stands for its interval midpoints.	required
search	Search or None	Ellipsoid of the volume form, from its radius, rotation and ratios; None for the plan form.	required
composite_length	float	c; default the median interval length of a Drillholes data. Required for other data with a search.	None
holes	array_like or str	Hole of each sample, or its column in data; the plan form then counts each hole once, at its nearest sample.	None
n	int or sequence of int	Neighbor ranks of the plan form, averaged; default 4 to 10.	None
hull	bool	Null targets outside the plan convex hull of data.	False

Returns:

Type	Description
ndarray	One spacing per target; NaN where the ellipsoid holds no composite, fewer than n + 1 samples or holes exist, or outside the hull.

Warns:

Type	Description
UserWarning	When most targets have fewer than 4 composites inside the ellipsoid: enlarge the search.

used in

• Drill-hole spacing from a learning curve
• outlines
• data spacing
• classification
• A coal seam in a lease