VTFIT: a microcomputer-based routine for fitting probability distribution functions to data

Richard A C Cooke, S. Mostaghimi, F. E. Woeste

Research output: Contribution to journalArticle

Abstract

A routine was developed to fit probability distribution functions to data, using the maximum likelihood method. The distributions that can be fitted include the Gaussian (normal), two-parameter log Gaussian (log normal), three-parameter log Gaussian, exponential, shifted exponential, beta, gamma, three-parameter gamma (Pearson type III), log gamma (log Pearson type III), inverted gamma (Pearson type V), Gumbel (extreme value type I) for minima, Gumbel (extreme value type I) for maxima, Frechet (extreme value type II) for minima, Frechet (extreme value type II) for maxima, three-parameter Frechet for maxima; Weibull (extreme value type III) for maxima, Weibull (extreme value type III) for minima, and three-parameter Weibull for minima distributions. Test statistics are provided for the chi-square, Kolmogorov-Smirnov, Kuiper, Cramer-von Mises, Anderson-Darling, and maximum likelihood goodness-of-fit tests. Graphs of the frequency histogram with the fitted probability distribution function superimposed, and of the empirical distribution function for the sample data, together with the fitted cumulative density function, are provided for visual assessment.

Original languageEnglish (US)
Pages (from-to)401-408
Number of pages8
JournalApplied Engineering in Agriculture
Volume9
Issue number4
StatePublished - Jul 1993
Externally publishedYes

ASJC Scopus subject areas

  • Agricultural and Biological Sciences (miscellaneous)

Fingerprint Dive into the research topics of 'VTFIT: a microcomputer-based routine for fitting probability distribution functions to data'. Together they form a unique fingerprint.

  • Cite this