GradientLat

DESCRIPTION

This module computes the gradient of data on a lattice of any type, up to three dimensions. It creates an output lattice with the same coordinates as input, with data being the computed gradient vectors.

The module uses central differencing for uniform and perimeter lattices, and NAG routines for curvilinear lattices. The NAG routines employ a quadratic Shepard method that handles duplicate points and coplanar points.

This module is meant as a replacement for Gradient and Gradient3D.

INPUTS

Port: Lattice In
Type: Lattice
Constraints: -D
Constraints: -vector
Constraints: byte..double
The input lattice.

WIDGETS

Port: Scale
Type: Dial
A scale factor to be applied to the output vector.

Port: Channel
Type: Slider
If the input data is itself vector, this selects the data channel.

OUTPUTS

Port: Gradient
Type: Lattice
Constraints: -D
Constraints: 1..3-vector
Constraints: float
The output lattice of vectors.

PROBLEMS

The curvilinear routines are considerably slower than the uniform and perimeter routines.

SEE ALSO

[ Documentation Home ]
© The Numerical Algorithms Group Ltd, Oxford UK. 1999