Least squares spline surface fitting program using B-splines. [LSB]

Technical Report · Mon May 01 00:00:00 EDT 1978

OSTI ID:6367378

Dickinson, R P; Carlson, R E

This report describes the mathematics used in a program for fitting, by least squares, a piecewise polynomial surface in two independent variables to a very large data base. The domain of the fit is assumed to lie within a rectangle, but the given (x,y) points may be irregularly spaced within the domain; that is, one does not assume that the (x,y) points lie on some sort of rectangular partition of the domain. Because of irregular spacing, often the data base will have areas of sparseness within the domain. These areas can cause severe stability problems when fitting. This report also describes methods to help overcome this stability problem. The surface of the fit is represented as a linear combination of products of B-spline basis functions, and the fit is accomplished by solving the normal equations. 4 figures.

• Research Organization: California Univ., Livermore (USA). Lawrence Livermore Lab.
• DOE Contract Number: W-7405-ENG-48
• OSTI ID: 6367378
• Report Number(s): UCID-17807
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
54 ENVIRONMENTAL SCIENCES
COMPUTER CODES
L CODES
LEAST SQUARE FIT
COMPUTER CALCULATIONS
MATRICES
OZONE
POLYNOMIALS
SPATIAL DISTRIBUTION
STABILITY
SURFACES
THREE-DIMENSIONAL CALCULATIONS
DISTRIBUTION
FUNCTIONS
MAXIMUM-LIKELIHOOD FIT
NUMERICAL SOLUTION
990200* - Mathematics & Computers
500200 - Environment
Atmospheric- Chemicals Monitoring & Transport- (-1989)