def kcluster(data, nvals):
data = data.astype(np.float)
from vlfwrap import KMeans
km = KMeans()
km.repetitions = 1
def kmeans(data, k):
km.cluster(data, k)
return km.quantize(data)
return cluster_with_automatic_k_selection(data, nvals, kmeans)
def kcluster_fixed(data, k):
data = data.astype(np.float)
from vlfwrap import KMeans
km = KMeans()
km.repetitions = 1
km.cluster(data, k)
qa = km.quantize(data)
nc = qa.max() + 1
# if nc != k and len(data) >= k:
# qa = np.arange(len(data))
# qa %= k
# nc = k
return qa, nc