Accession Number : AD0687276

Title :   A GENERAL NON-LINEAR LEAST SQUARES DATA-FITTING PROGRAM,

Corporate Author : ARMY NUCLEAR DEFENSE LAB EDGEWOOD ARSENAL MD

Personal Author(s) : Vogt,David ; Buhl,A. R.

Report Date : APR 1969

Pagination or Media Count : 53

Abstract : The paper describes a general non-linear Least Squares Program (LSP). LSP will accept as many as 200 data values and will fit these data with any equation that conforms to the following criteria: is continuously differentiable; is a function of 5 or less independent variables; is linear or non-linear; and has twenty, or less, fitting parameters. In accordance with certain convergence criteria, the program calculates the values of the fitting parameters for which the given equation best fits the input data. The associated variances are calculated per data point and per fitting parameter. The program is written in FORTRAN IV and requires three tape drives and a 16K memory. (Author)

Descriptors :   (*LEAST SQUARES METHOD, COMPUTER PROGRAMS), CURVE FITTING, EXPERIMENTAL DATA, SUBROUTINES

Subject Categories : Statistics and Probability
      Computer Programming and Software

Distribution Statement : APPROVED FOR PUBLIC RELEASE