Chapter Contents

Previous

Next

Introduction

The G4GRID procedure creates a SAS data set to be used with the SAS/SPECTRAVIEW three-dimensional visualization software. The procedure enhances SAS/SPECTRAVIEW software's ability to create two- and three-dimensional graphic images from data.

PROC G4GRID allows you to interpolate three-dimensional data contained in a SAS data set. While PROC G4GRID is not required to load a data set into SAS/SPECTRAVIEW, the procedure does provide advantages:

• It enables the loading of data sets that could not otherwise be loaded due to memory constraints. SAS/SPECTRAVIEW creates a three-dimensional volume grid with values for each unique x,y,z coordinate contained in the data set. For sparse data sets, this can result in data points, many of which may be missing. By using PROC G4GRID prior to loading the data set, you can replace filler points with interpolated values or resize the data set as required.

• It creates larger or smaller resolutions of a data set to enhance or facilitate your view of the data. Using PROC G4GRID, you can create a data set of almost any desired resolution for use in SAS/SPECTRAVIEW.

• It creates more meaningful volume data from randomly sampled data points.

PROC G4GRID works by interpolating randomly distributed points in three dimensions and by creating a volume grid with data points at each grid location. The procedure uses a Gaussian interpolation function that varies as the inverse of the square of the distance from surrounding data points. While PROC G4GRID is similar in syntax and function to the G3GRID procedure, G4GRID uses a different interpolation method and does not render the same values as G3GRID.

You can run the procedure using either of these methods:

• PROC G4GRID syntax

• graphical interface.

Data Considerations

Each data set processed by the G4GRID procedure must contain at least four numeric variables:

• three axis variables

• one or more response variables.

If you specify more than one response variable, the procedure performs a separate analysis and produces interpolated values for each response variable. Unlike the G3GRID procedure, the G4GRID procedure does not require that the input data coordinate values be unique. It uses all coordinate values that fall within the grid volume.

Terminology

The following terms are specific to the G4GRID procedure:

coordinates	the values representing the location of a data point along the X, Y, Z axes.
interpolate	to estimate values between two or more known values.
response	a value representing a measured response associated with each data point in the volume.
variables	the SAS data set variables whose values are plotted on the X, Y, and Z axes.

Chapter Contents

Previous

Next

Top of Page

Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.