org.das2.qds.util

Class Griddata

java.lang.Object

org.das2.qds.util.Griddata

public class Griddata
extends java.lang.Object

Codes for interpolating from irregular grids to regular grids. For these codes we use "trigulations" which are datasets that contain connections of points in other datasets. These are rank 2 datasets tri[n,3] where the 3 points are indeces of other datasets. Triangulations are generalized, where the typical case is a set of 3-node triangles. However, 2-point "triangles" can be used for linear interpolation, 4-point squares can be used for bilinear interpolation, and 4-points for 3-D triangle interpolation. DO NOT USE THIS CLASS, IT IS STILL UNDER DEVELOPMENT! This is barely tested for the nfree=1-D case!

Constructor Summary

Constructors

Constructor and Description
Griddata()

Method Summary

Modifier and Type	Method and Description
static QDataSet	griddata(QDataSet xx, QDataSet yy, QDataSet buck)
static QDataSet	griddata(QDataSet triangles, QDataSet itriangle, QDataSet weights, QDataSet zbuck) Perform the interpolation for arbitrary sets of nvert points.
static QDataSet	triangulate(QDataSet buck) return a rank 2 dataset [n,nvert] with the "triangles" connecting the buckshot data.
static QDataSet	triLocate(QDataSet buck, QDataSet tri, QDataSet grid) for each element of grid, identify the matching triangle.
static QDataSet	weights(QDataSet buck, QDataSet tri, QDataSet grid, QDataSet itri) return the weights for each point of the grid.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Griddata

public Griddata()

Method Detail

griddata

public static QDataSet griddata(QDataSet triangles,
                                QDataSet itriangle,
                                QDataSet weights,
                                QDataSet zbuck)

Perform the interpolation for arbitrary sets of nvert points.

Parameters:

triangles - triangles mesh, presently tri[n,3] but may soon be tri[n,4] as well.

itriangle - rank n mesh identifying the triangle to use

weights - rank n+1 mesh (e.g. weights[m,3]) with a weight for each triangle node.

zbuck - dependent Z values for each point.

Returns:

dataset with the same geometry as itriangle.

griddata

public static QDataSet griddata(QDataSet xx,
                                QDataSet yy,
                                QDataSet buck)

triangulate

public static QDataSet triangulate(QDataSet buck)

return a rank 2 dataset [n,nvert] with the "triangles" connecting the buckshot data.

Parameters:

buck - dataset[m,nvert-1].

Returns:

tri[n,3]

triLocate

public static QDataSet triLocate(QDataSet buck,
                                 QDataSet tri,
                                 QDataSet grid)

for each element of grid, identify the matching triangle.

Parameters:

buck - the data (e.g. buck[o,nfree])

tri - the triangulation (e.g. tri[n,free+1])

grid - the values to locate (e.g. grid[m,nfree])

Returns:

the locations itri[o]

weights

public static QDataSet weights(QDataSet buck,
                               QDataSet tri,
                               QDataSet grid,
                               QDataSet itri)

return the weights for each point of the grid. Let nfree be the number of independent data dimensions.

Parameters:

buck - the data (e.g. buck[o,nfree])

tri - the triangulation (e.g. tri[n,nfree+1])

grid - rank 2 dataset (e.g. grid[m,nfree])

itri - the triangle to use for each point of grid. (e.g. itri[m])

Returns:

rank 2 (e.g. weights[m,2])