A wide range of mathematical transforms can be applied using this option.
The selected operation is applied to all elements of the data set.
The data set may be observations, interpolated surface, cross-validation
errors, or a previously
transformed variable.
Linear transform
Change the units or rescale the data by specifying a constant and slope.
Nonlinear transforms
Used mainly to reduce variance in the data. The inverse transform is
applied automatically following interpolation, if desired. Such an
option appears when the Begin Interpolation is selected.
The available nonlinear transforms are:
Log [ln(x)]
Exponential [exp(x)]
Power [x^{p}]
Box-Cox [(x^{p})/p]
Logistic [ln[x/(1-x)]]
Angular [arcsin(sqrt(x))]
Trigonometric [sin(x) or cos(x)]
Average
Use this option to take the average of several data files. Each file
must have
the same dimension. This option can be used to create a monthly or
annual average of a quantity prior to (or subsequent to)
interpolation.
Formula
Using an equation editor, a user can create new variables and perform the
interpolation directly
on these new variables. If desired, the operations creating the variable
are automatically
inverted following interpolation, to restore physical meaning to the
interpolation
variable.
Using this option, a variogram is fitted to the data set.
The domain of a dataset can be limited to a rectangle in the
latitude-longitude plane. To use this option, the user enters a bounding
pair of latitudes and longitudes.
The values in a dataset can be constrained to lie within specified minimum
and maximum values.