NAME:  	Fisher Iris Data (IRIS.DAT)
KEYWORDS:  Discriminant Function Analysis
SIZE:  150 observations, 5 variables

SOURCE:  Fisher, R.A. (1936), The use of multiple measurements in taxonomic
         problems.  Annals of Eugenics, 7:179-188. 


DESCRIPTIVE ABSTRACT:

This classic data set originally appeared in R.A. Fisher's 1936 paper in
which discriminant function analysis was first developed.  The iris data
contain measurements of sepal length and width, and petal length and
width on samples of 50 irises from each of three species.  Each data
record contains the four measurments for one flower, plus a code for
species: setosa, versicolor, and virginica. 


LIST OF VARIABLES:

Variable		Description
_________________________________________________________________

V1	     Species (Iris Type), 1 = setosa,  2 = versicolor, 
        			  3 = virginica.

V2	     Sepal Length in centimeters.

V3	     Sepal Width in centimeters.

V4	     Petal Length in centimeters.

V5	     Petal Width in centimeters. 
__________________________________________________________________


PEDAGOGICAL NOTES:

It is of interest to determine which linear combination of these variables 
best discriminates between irises from different species, and how much
overlap between different species remains after these characteristics
have been used for classification.