TY - RPRT
TI - Fitting of two and three variate polynomials from experimental data through the least squares method; Ajuste de polinomios en dos y tres variables independientes por el mtodo de mnimos cuadrados
AB - Obtaining polynomial fittings from observational data in two and three dimensions is an interesting and practical task. Such an arduous problem suggests the development of an automatic code. The main novelty we provide lies in the generalization of the classical least squares method in three FORTRAN 77 programs usable in any sampling problem. Furthermore, we introduce the orthogonal 2D-Legendre function in the fitting process. These FORTRAN 77 programs are equipped with the options to calculate the approximation quality standard indicators, obviously generalized to two and three dimensions (correlation nonlinear factor, confidence intervals, cuadratic mean error, and so on). The aim of this paper is to rectify the absence of fitting algorithms for more than one independent variable in mathematical libraries.
AU - Sanchez-Miro, J J
AU - Sanz-Martin, J C
KW - 99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
KW - LEGENDRE POLYNOMIALS
KW - THREE-DIMENSIONAL CALCULATIONS
KW - FORTRAN
KW - COMPUTER CODES
KW - TWO-DIMENSIONAL CALCULATIONS
KW - LEAST SQUARE FIT
KW - MATHEMATICS
KW - 990200
KW - MATHEMATICS AND COMPUTERS
DO -
UR -
PB -
CY - Spain
PY - 1994
DA - 1994-12-31
LA - Spanish
J2 -
C1 - Centro de Investigaciones Energeticas, Medioambientales y Tecnologicas (CIEMAT), Madrid (Spain)
C2 -
ER -