At present, the methods available are as follows.
Fourier Analysis (default) The input data is assumed to be an sampled set of points from a function which can be reconstructed using 2D fourier analysis. Given the input values, the frequency components of the estimated function are computed. The values at each output grid point are computed using this function.
Inverse Distance Method At each output grid point, all points that lie within a user-defined search radius are averaged together using an inverse distance formula. The user is provided with an exponent which is applied to the denomimator of this forumula. Increasing the value of the exponent will tend to cause points farther away to have less influence. Decreasing the value of the exponent will tend to make the surface a bit smoother.
Common problems:
Search Radius Too Small The search radius must be large enough so as to have some overlap with adjacent output grid points, and large enough to include input data. In areas of sparse input data, often the output grid will have it's value set to "UNDEFINED", as no input points are close enough to influence this grid point. The two ways to work around this are to use a larger search radius (the radius is in the same space as the input data, by the way), or to use a different method which works better with sparse data, such as Fourier Analysis.
Results of Fourier Analysis too Extreme On occasion, this method will produce values which vary wildly from the input, resulting in surfaces which have marked peaks or valleys, seemingly going off into infinity. This is produced by extrapolation into a region where there is no data, based upon a partial derivative computed at two data points close together. The only way to work around this is to include more points in the surface computation, or to use a method which doesn't use partial derivatives, such as the Inverse Distance Method.
COMING SOON: Bivariate Analysis (Fall 1994)
The extents of the output grid, by default, are computed for you. If this isn't acceptable, then set the min/max values for the grid yourself.
Mutually Exclusive Group; if desired, specify ONE of:
ALL OF the Mutually Inclusive Group:
AND
AND
AND
ALL OF the Mutually Inclusive Group:
AND
Mutually Exclusive Group; if desired, specify ONE of:
OR
ALL OF the Mutually Inclusive Group:
AND
AND
AND
AND