Part of a revelation I’ve had this past couple of months relates to how different programs are good at doing different things.

For instance, R is very intuitive and fast when it comes to doing classifications and projections. Here is how you would implement a Naive Bayes and K-NN in R.

**naive Bayes in R :**

make sure the columns for train and test have the same heading.

>

` train<-read.csv("train_7dim.`

csv")

```
> test<-read.csv("test_7dim.csv"
```)
> dft <- data.frame(cbind(train,label))
> head(dft)

```
> dft$X0 <- as.factor(dft$X0)
> m<-naiveBayes(X0~.,dft)
> table(predict=predict(m, test))
```

**k-nn in R**:

>

`knn(train, test, label, k = 1, l = 0, prob = FALSE, use.all = TRUE)`