#############DISCRIMINANT FUNCTION ANALYSIS
> library(MASS, pos=16)


> discr1 <- lda (Species ~ Sepal.Length + Sepal.Width + Petal.Length, data = iris)

> discr1
Call:
lda(Species ~ Sepal.Length + Sepal.Width + Petal.Length, data = iris)

Prior probabilities of groups:
    setosa versicolor  virginica 
 0.3333333  0.3333333  0.3333333 

Group means:
           Sepal.Length Sepal.Width Petal.Length
setosa            5.006       3.428        1.462
versicolor        5.936       2.770        4.260
virginica         6.588       2.974        5.552

Coefficients of linear discriminants:
                    LD1        LD2
Sepal.Length -1.2292136 -0.6402838
Sepal.Width  -0.9707919  3.0792137
Petal.Length  3.2507528  0.6580622

Proportion of trace:
   LD1    LD2 
0.9919 0.0081 

> plot (discr1)

> plot (discr1, col = as.integer(iris$Species))

> library(psych, pos=17)

> pairs.panels(iris[1:4])

### PERCENTAGE OF CORRECT CLASSIFICATION

> correct <- predict (discr1, iris)$class

> iris.table <- table(predicted = correct, Actual = iris$Species)

> iris.table
            Actual
predicted    setosa versicolor virginica
  setosa         50          0         0
  versicolor      0         48         3
  virginica       0          2        47



#### HISTORGAMS FOR LD1

> iris.values <- predict (discr1)


> ldahist(iris.values$x[,1], g = iris$Species)

> ldahist(iris.values$x[,2], g = iris$Species)

### CROSS VALIDATION

> attach(iris)

> iris.train1 <- sample (1:150, 70)

> table (iris$Species[iris.train1])

    setosa versicolor  virginica 
        21         23         26 

> discr2 <- lda (Species ~ Sepal.Length + Sepal.Width + Petal.Length, data = iris, prior = c(1/3, 1/3, 1/3), subset = iris.train1)

> predict (discr2, iris[-iris.train1, ])$class
 [1] setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa    
[12] setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa    
[23] setosa     setosa     setosa     setosa     setosa     setosa     setosa     versicolor versicolor versicolor versicolor
[34] versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor
[45] versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor
[56] versicolor virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica 
[67] virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  virginica  versicolor virginica 
[78] virginica  virginica  virginica 
Levels: setosa versicolor virginica

########################################################
################SCRIPT#################################

#############DISCRIMINANT FUNCTION ANALYSIS
library(MASS, pos=16)
discr1 <- lda (Species ~ Sepal.Length + Sepal.Width + Petal.Length, data = iris)
discr1
plot (discr1)
plot (discr1, col = as.integer(iris$Species))

library(psych, pos=17)
pairs.panels(iris[1:4])

### PERCENTAGE OF CORRECT CLASSIFICATION
correct <- predict (discr1, iris)$class
iris.table <- table(predicted = correct, Actual = iris$Species)
iris.table

#### HISTORGAMS FOR LD1 and LD2
iris.values <- predict (discr1)
ldahist(iris.values$x[,1], g = iris$Species)
ldahist(iris.values$x[,2], g = iris$Species)


### CROSS VALIDATION
attach(iris)
iris.train1 <- sample (1:150, 70)
table (iris$Species[iris.train1])
discr2 <- lda (Species ~ Sepal.Length + Sepal.Width + Petal.Length, data = iris, prior = c(1/3, 1/3, 1/3), subset = iris.train1)
predict (discr2, iris[-iris.train1, ])$class