The keyword which defines the type of transformation to be applied to the Model. This keyword must be chosen among the following list:

• none : to undefine any Model transformation

• conv : to apply a convolution to the Model

• anam : to apply the Discrete Anamorphosis transform to the Model

• tape : to apply a Tapering transformation to the Model

Only one Model transformation can be applied at a time. Each Model transformation requires a set of relative parameters: their name use the transformation name as a radix.

Type of convolution function. It must be selected from the following list:

• 1 : Uniform

• 2 : Exponential

• 3 : Gaussian

• 4 : Cardinal Sine

Direction in which the convolution must be applied. It must be selected from the following list:

• 1 : Along the X-axis only

• 2 : Along the Y-axis only

• 3 : Along the Z-axis only

• 4 : In the XY plane

• 5 : In the XYZ volume

The convolution is performed using a numerical integration. This is performed through a discretization. The current argument provides the number of discretization steps along each direction where the convolution takes place. Beware that choosing 'ndisc=10' in the case where 'idir=5' leads to 1000 discretization steps for each covariance calculation.

This parameter defines the (apparent) range of the convolution along each direction where the convolution is defined.

This argument refers to the anam-class object which contains the (block) Discrete Anamorphosis definition.

In the Discrete Disjunctive Kriging, the aim is estimated the Discretized Variable. This is performed via the (disjunctive) estimation of the (independent) factors. However, the user may want to estimate ONE factor individually: then it suffices to specify the rank of this factor in this argument (starting from 1)

in the case of Discrete Indicator Residuals Anamorphosis, the list of Models (for each active Residuals) must be provided. The number of Models in this list must match the number of active residuals provided by argument 'ncleff'.

This argument gives the type of covariance or variogram to be computed:

• P : the variogram or the covariance of the punctual variable

• B : the variogram or the covariance of the block variable

• C : the cross-variogram or the cross-covariance between the point and the block variables

In the case of Tapering Model calculation, the user must choose the tapering function (which is a covariance with compact support) in the following list:

• 1 : Spherical

• 2 : Cubic

• 3 : Triangle

• 4 : Penta Model

• 5 : Storkey Model (1-D)

• 6 : Wendland1

• 7 : Wendland2

Apparent range of the tapering function.