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] aclj_v1_0.gz(52 Kbytes)
Manuscript Title: Programs for signal recovery from noisy data using the maximum likelihood principle. II. Program implementation.
Authors: V.I. Gelfgat, E.L. Kosarev, E.R. Podolyak
Program title: MLG8, MLU8, MLP8
Catalogue identifier: ACLJ_v1_0
Distribution format: gz
Journal reference: Comput. Phys. Commun. 74(1993)349
Programming language: Fortran.
Computer: IBM PS/2 mod.50.
Operating system: PC DOS 3.30.
RAM: 200 K words
Word size: 16
Keywords: Ill-posed problems, Maximum likelihood, Steepest descent, General purpose, Quadrature.
Classification: 4.11.

Nature of problem:
A program package for non-negative signal recovery from noisy experimental data distorted by the measuring device is presented. Noise may have a Gaussian, binomial or Poissonian distribution at each experimental point. Three examples are given which demonstrate the usage of the package: a) reduction to an arbitrary instrumental function, which depends only on the difference of its arguments (this is the typical spectroscopic problem); b) the expansion of an exponentially decaying curve which has a continuous spectrum of decrements (this problem has applications in time dependent fluorescence, nuclear physics etc.); and c) the ultrasoft X-Ray spectrum recovery from the absorption measurements. These programs are based on the Maximum Likelihood principle which allows the maximum value of resolution enhancement according to Shannon's information theory to be achieved.