O. Conrad (c) 2008
This module approximates irregular scalar 2D data in specified points using C1-continuous bivariate cubic spline.
Minimal Number of Points: minimal number of points locally involved in spline calculation (normally = 3)
Maximal Number of Points:npmax: maximal number of points locally involved in spline calculation (required > 10, recommended 20 < npmax < 60)
Tolerance: relative tolerance multiple in fitting spline coefficients: the higher this value, the higher degree of the locally fitted spline (recommended 80 < k < 200)
Points per square: average number of points per square (increase if the point distribution is strongly non-uniform to get larger cells)
Author: Pavel Sakov, CSIRO Marine Research
Purpose: 2D data approximation with bivariate C1 cubic spline. A set of library functions + standalone utility.
Description: See J. Haber, F. Zeilfelder, O.Davydov and H.-P. Seidel, Smooth approximation and rendering of large scattered data sets, in 'Proceedings of IEEE Visualization 2001' (Th.Ertl, K.Joy and A.Varshney, Eds.), pp.341-347, 571, IEEE Computer Society, 2001.
www.uni-giessen.de/www-Numerische-Mathematik/davydov/VIS2001.ps.gz
www.math.uni-mannheim.de/~lsmath4/paper/VIS2001.pdf.gz