Clustering and Classification methods for Biologists


MMU logo

Discriminant Analysis

LTSN Bioscience logo

Page Outline

 

Search

[ Yahoo! ] options

Iris Data : Discriminant Analysis 1 (4 pedictors)

These are the results of a discriminant analysis using all four predictors

Summary of Canonical Discriminant Functions

Eigenvalues
Function Eigenvalue % of Variance Cumulative % Canonical Correlation
1 32.192 99.1 99.1 .985
2 .285 .9 100.0 .471

 

Wilks' Lambda
Test of Function(s) Wilks' Lambda Chi-square df Sig.
1 through 2 .023 546.115 8 .000
2 .778 36.530 3 .000

 

top

Standardized Canonical Discriminant Function Coefficients

Function
1 2
SEPAL_L -0.427 0.012
SEPAL_W -0.521 0.735
PETAL_L 0.947 -0.401
PETAL_W 0.575 0.581

 

Structure Matrix

Function
1 2
PETAL_L 0.706 0.168
SEPAL_W -0.119 0.864
PETAL_W 0.633 0.737
SEPAL_L 0.223 0.311

 

Functions at Group Centroids
  Function
SPECIES 1 2
I.setosa -7.608 .215
I.versicolor 1.825 -.728
I.virginica 5.783 .513

 

top

Classification Statistics

Prior Probabilities for Groups
  Prior Cases Used in Analysis
SPECIES   Unweighted
I.setosa .333 50 50.000
I.versicolor .333 50 50.000
I.virginica .333 50 50.000
Total 1.000 150

 

 
Classification Results
Predicted Group Membership Total
    SPECIES I.setosa I.versicolor I.virginica
Original Count I.setosa 50 0 0 50
I.versicolor 0 48 2 50
I.virginica 0 1 49 50
% I.setosa 100.0 .0 .0 100
I.versicolor .0 96.0 4.0 100
I.virginica .0 2.0 98.0 100
Cross-validated Count I.setosa 50 0 0 50
I.versicolor 0 48 2 50
I.virginica 0 1 49 50
% I.setosa 100.0 .0 .0 100
I.versicolor .0 96.0 4.0 100
I.virginica .0 2.0 98.0 100
top