Kansas Geological Survey, Computer Contributions 47, originally published in 1970
University of Uppsala
Originally published in 1970 as Kansas Geological Survey Computer Contributions 47.
This FORTRAN IV program for canonical variates for a maximum of 30 variables performs several functions: (1) basic statistics for each sample are computed and the covariance matrices tested for homogeneity; (2) the generalized determinantal equation is solved and eigenvalues tested for significance; (3) the mean vectors are tested in a one-way multivariate analysis of variance for significant difference; and (4) the first two elements of the transformed mean vectors and transformed observations are plotted.
Application of the program is illustrated by two examples: observations on Upper Cretaceous echinoids from England and matrix input data from an anthropomorphic problem.
The FORTRAN IV program accomplishes the calculations required for a canonical variate analysis including the associated multivariate analysis of variance. The theory is well known and may be consulted in many textbooks on multivariate analysis.
Canonical variate analysis, referred to as "canvar", is a logical extension of the analysis of two multivariate samples for which a comprehensive FORTRAN IV program was published recently by Reyment, Ramdén, and Wahlstedt.
Two problems are used as examples of application of the program. Data for the first example are on Upper Cretaceous echinoids from England (Nicholls, 1959). Data for the second example are from an anthropomorphic problem (Talbot and Mulhall, 1962). The examples are given in the Appendix.
The original work for this paper was done while R.A.R. was Visiting Research Scientist with the Kansas Geological Survey in 1966-1967. The work was supported by Computing Grant 104104 at the University of Uppsala and Swedish Natural Science Council Research Contract 2320-29.
Read the PDF version (3.8 MB)
Kansas Geological Survey
Placed on web Sept. 3, 2019; originally published 1970.
Comments to webadmin@kgs.ku.edu
The URL for this page is http://www.kgs.ku.edu/Publications/Bulletins/CC/47/index.html