dissemination builtin

dissemination(madogram, variogram)

Degree of dissemination per lag, √π · M(h) / √γ(h).

M is the madogram and γ the classical variogram of the same values, lags and directions. Gaussian increments give 1. Below 1, a few large differences carry γ while most pairs differ little, as when high values sit in isolated samples; above 1, differences are more even than Gaussian.

Parameters:

Name	Type	Description	Default
madogram	ExperimentalVariogram	From estimator="madogram".	required
variogram	ExperimentalVariogram	From estimator="matheron" with the same arguments otherwise.	required

Returns:

Type	Description
ndarray	One value per lag, NaN where γ is 0.

used in

• madogram