Machine Learning With Go

Generating clusters with k-means

Now let's get our hands dirty by actually applying k-means clustering to the delivery driver data. To utilize github.com/mash/gokmeans, we first need to create a slice of gokmeans.Node values, which will be input into the clustering:

// Open the driver dataset file.f, err := os.Open("fleet_data.csv")if err != nil {    log.Fatal(err)}defer f.Close()        // Create a new CSV reader. r := csv.NewReader(f) r.FieldsPerRecord = 3 // Initialize a slice of gokmeans.Node's to // hold our input data. var data []gokmeans.Node // Loop over the records creating our slice of // gokmeans.Node's. for { // Read in our record and check for errors. record, err := r.Read() if err == io.EOF { break } if err != nil { log.Fatal(err) } // ...