Go implementations of the DBSCAN and k-means clustering algorithms, the k-d tree spatial index and some functions to generate and visualize sample data.

Example Output

Below are two examples that show the algorithms in action. In the output, clusters are indicated by colors; points shown in gray were marked as noise.

The first example is 150 points clustered by k-means:

k-means example output

and DBSCAN:

DBSCAN example output

The second example includes a cluster of points distributed in a non-convex shape. K-means doesn't recognize it:

k-means example output

but DBSCAN does:

DBSCAN example output