The notebook has the descriptions of results and what we are doing.

Overall I tested the different between Kmean and Kmean++ from 1 to 10 clusters. Also showing the graph of the change in data after each clusters. We can see that after 3 clusters we get that sick elbow.

After that we each of point in the 3 clusters to see of they are a good fit or not. This is shown in the shilhouette ceofficient graph. The points are pretty good.