Kamaljeet Singh (19103052)

Loading the dataset

In [4]: df<-datasets::iris

Checking the dataset

In [6]: head(df,10)

Sepal.Length Sepal.Width Petal.Length Petal.Width Species

5.1 3.5 1.4 0.2 setosa

4.9 3.0 1.4 0.2 setosa

4.7 3.2 1.3 0.2 setosa

4.6 3.1 1.5 0.2 setosa

5.0 3.6 1.4 0.2 setosa

5.4 3.9 1.7 0.4 setosa

4.6 3.4 1.4 0.3 setosa

5.0 3.4 1.5 0.2 setosa

4.4 2.9 1.4 0.2 setosa

4.9 3.1 1.5 0.1 setosa

Function to split the dataset into train and test set

Ratio--> 0.3 means 30% rows used as test set

In [7]: train_test_split<-function(df,test_ratio){

set.seed(1)

split<- sample(c(rep(0, (1-test_ratio) * nrow(df)), rep(1, test_ratio * nrow(df

))))

return(split)

Function to calculate Euclidean Distance

In [8]: euclidean_dist<-function(train_row,test_row){
dist<-sum((test_row[1:4]-train_row[1:4])**2)

return(dist)

Function to calculate the euclidean distance of one test instance wrt all the training instances

 In [9]: calc_distances<-function(train_df,test_row){

distances<-c(euclidean_dist(train_df[1,],test_row),euclidean_dist(train_df[2,],te
st_row))

classes<-c(as.numeric(train_df[1,][5]),as.numeric(train_df[2,][5]))

n<-nrow(train_df)

for(i in 3:n){

curr_dist=euclidean_dist(train_df[i,],test_row)

curr_class=as.numeric(train_df[i,][5])

distances<-append(distances,curr_dist)

classes<-append(classes,curr_class)

distances=cbind(data.frame(distances), data.frame(classes))

return(distances)

Function to sort the distances and then selecting the majority class with minimum distance

In [11]: predict_class<-function(train_df,test_row,k){
distances<-calc_distances(train_df,test_row)

distances<-distances[order(distances[,1]),]

dist<-table(distances[1:k,2])

class=names(dist)[which.max(dist)]

return(class)

Function to perform knn for all test instances

In [12]: knn<-function(train_df,test_df,k){

n<-nrow(test_df)

predictions<-list()

for(i in 1:n){

curr_pred<-predict_class(train_df,test_df[i,],k)

predictions[length(predictions)+1]<-curr_pred

return(predictions)

Function to calculate accuracy

In [13]: calc_accuracy<-function(test_classes,pred){

correct<-sum(test_classes==pred)

return(correct/length(test_classes))

Creating the train and test sets form the performed split

In [14]: split<-train_test_split(df,0.3)

train_df<-df[split==0,]

test_df<-df[split==1,]

Checking the train and test sets

 In [15]: head(train_df,5)

Sepal.Length Sepal.Width Petal.Length Petal.Width Species

1 5.1 3.5 1.4 0.2 setosa

3 4.7 3.2 1.3 0.2 setosa

4 4.6 3.1 1.5 0.2 setosa

5 5.0 3.6 1.4 0.2 setosa

6 5.4 3.9 1.7 0.4 setosa

In [16]: head(test_df,5)

Sepal.Length Sepal.Width Petal.Length Petal.Width Species

2 4.9 3.0 1.4 0.2 setosa

8 5.0 3.4 1.5 0.2 setosa

15 5.8 4.0 1.2 0.2 setosa

17 5.4 3.9 1.3 0.4 setosa

18 5.1 3.5 1.4 0.3 setosa

Funtion to change the classes from numeric to string in the predicted values

In [17]: correct_classes<-function(pred){

n<-length(pred)

for(i in 1:n){

if(pred[i]=='1')

pred[i]='setosa'

else if(pred[i]=='2')

pred[i]='versicolor'

else

pred[i]='virginica'

return(pred)

Performing KNN with k-value as 5

In [18]: pred<-knn(train_df,test_df,5)

pred=correct_classes(pred)

 In [19]: pred

1. 'setosa'
2. 'setosa'
3. 'setosa'
4. 'setosa'
5. 'setosa'
6. 'setosa'
7. 'setosa'
8. 'setosa'
9. 'setosa'
10. 'setosa'
11. 'setosa'
12. 'setosa'
13. 'setosa'
14. 'setosa'
15. 'versicolor'
16. 'versicolor'
17. 'versicolor'
18. 'versicolor'
19. 'versicolor'
20. 'versicolor'
21. 'versicolor'
22. 'versicolor'
23. 'versicolor'
24. 'virginica'
25. 'versicolor'
26. 'virginica'
27. 'versicolor'
28. 'versicolor'
29. 'versicolor'
30. 'versicolor'
31. 'versicolor'
32. 'versicolor'
33. 'virginica'
34. 'versicolor'
35. 'virginica'
36. 'virginica'
37. 'virginica'
38. 'virginica'
39. 'virginica'
40. 'virginica'
41. 'virginica'
42. 'virginica'
43. 'virginica'
44. 'virginica'
45. 'virginica'

In [20]: calc_accuracy(test_df$Species,pred)

0.933333333333333

Perform knn for odd k values from 3 to 31 and then comparing accuracies

In [ ]: k_values<-seq(3,31,by=2)

accuracies<-list()

for(i in k_values){

preds=knn(train_df,test_df,i)

preds=correct_classes(preds)

accuracy=calc_accuracy(test_df$Species,preds)

accuracies<-append(accuracies,accuracy)

accuracies

In [ ]: