Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] abch_v1_0.gz(19 Kbytes)
Manuscript Title: Multivariate least squares fitting program using modified Gram- Schmidt transformations.
Authors: M.J. O'Connell
Program title: GSFIT
Catalogue identifier: ABCH_v1_0
Distribution format: gz
Journal reference: Comput. Phys. Commun. 8(1974)56
Programming language: Fortran.
Computer: IBM 370/195.
Operating system: IBM OS/HASP.
RAM: 37K words
Word size: 32
Peripherals: magnetic tape.
Keywords: General purpose, Nuclear physics, High energy, Least squares, Fit, Modified gram-schmidt Transformations, Orthogonal basis vectors, Chebyshev polynomials, Legendre polynomials, Multivariate analysis, General experiment.
Classification: 4.9, 17.4.

Nature of problem:
Fitting a dependent variable in terms of functions of many independent variables, e.g. finding the momentum of a charged particle from observation of its trajectory in magnetic field.

Solution method:
Residual sum of square is minimized using modified Gram-Schmidt o orthogonal basis vectors.

None in principle; coding arranged for 6 independent variables, 250 observations, up to 1000 trial functions.

Unusual features:
User has considerable control over the functions to be included, with the facility to reject marginally contributing terms.

Running time:
15 s on IBM 360/195.