data(iris)
set.seed(9850)
#Create the random rows 
gp<- runif(nrow(iris))   #runif will create the uniform distributions,they are between 0 and 1
iris1<- iris[order(gp),]
#Normalize each coloum of variables, we could use 0~1 function or z-scores,we use 0~1 here
normalize<- function(x){
  return((x-min(x))/(max(x)-min(x)))
}
normalize(c(1,2,3,4,5))

## [1] 0.00 0.25 0.50 0.75 1.00

iris_n<- as.data.frame(lapply(iris1[,c(1,2,3,4)],normalize))
#Create the training dataset and testing dataset
iris_train<- iris_n[1:129,]
iris_test<- iris_n[130:150,]
iris_train_target<- iris1[1:129,5]
iris_test_target<- iris1[130:150,5]
require(class)

## Loading required package: class

k=sqrt(nrow(iris1))
m1<- knn(train=iris_train,test=iris_test,cl=iris_train_target,k=k)
m1

##  [1] versicolor setosa     virginica  virginica  virginica  setosa    
##  [7] virginica  versicolor virginica  setosa     setosa     virginica 
## [13] setosa     virginica  virginica  virginica  setosa     virginica 
## [19] virginica  versicolor setosa    
## Levels: setosa versicolor virginica

table(iris_test_target,m1)  #Only 1 wrong, it is a pretty good result

##                 m1
## iris_test_target setosa versicolor virginica
##       setosa          7          0         0
##       versicolor      0          3         2
##       virginica       0          0         9