compare builtin

compare(estimate, truth, *, data=None)

Mean distance from each target to its n nearest drill holes, each hole at its nearest sample, for classification by drill spacing.

Parameters:

Name	Type	Description	Default
targets	(array_like, PointSet or BlockModel)	Locations to measure from.	required
coords	array_like	Sample coordinates, (n, 2) or (n, 3).	required
holes	array_like or str	Hole id of each sample, or the column of coords holding them; samples of one hole count once.	required
n	int or sequence of int	Number of holes averaged; one per class.	required
search	Search	Its ellipsoid sets the distance (meters along the major axis), and samples beyond its radius are ignored. Isotropic and unbounded if omitted.	required
domains	tuple of array_like	(target_domains, sample_domains): each target only sees samples of its own domain.	required
domain_column	str or tuple of str	The column of targets and coords holding their domains, or one name for each; instead of domains.	required

Returns:

Type	Description
ndarray	Distances, inf where fewer than n holes are in reach; shape (targets, len(n)) when n is a sequence.
How well `estimate` matches `truth`, e.g. an estimate against an
exhaustive grid; pairs with a missing value on either side are skipped.

Parameters:

Name	Type	Description	Default
estimate	array_like or str	One value per row, or column names of data.	required
truth	array_like or str	One value per row, or column names of data.	required
data	container	Holds the columns estimate and truth name.	None

Returns:

Type	Description
dict	n pairs compared, mean_error and rmse of estimate − truth, correlation, and slope of the regression of truth on estimate.

used in

• Kriging
• Checking a model
• Compare two estimates of one deposit
• quick tour
• imputation
• spatial imputation
• intrinsic coregionalization
• local variogram parameters