def compute_silhouette_block_scales(X, ward, scales, metric_measure, verbose=False):
if verbose == True:
print "Computing silhouette score :"
print "Scales : ", scales
sil = np.zeros(len(scales))
for i in np.arange(len(scales)):
height_cut = scales[i]
if verbose:
print ".....scales: ", height_cut
num_clusters = len(ward.cut(height_cut))
label = _hc_cut(num_clusters, ward.children_, ward.n_leaves_)
iterations = 1
temp = np.zeros(iterations)
for t in range(iterations):
st = cpu_time()
score = silhouette_score_block(X, label, metric=metric_measure, sample_size=5000, n_jobs=2)
t = cpu_time() - st
if verbose:
print "......time", t
temp[t] = score
sil[i] = temp[:].mean(0)
if verbose:
print ".....score: ", score
mean = sil[:].mean(0)
std = sil[:].std(0)
return mean, std
def compute_silhouette_score(X, tree, metric_measure):
'''
n : sample sizes |X|
num of clusters, k = [1..n]
for each value of k
P_k: partition of X having k cluster (based on the maximum distance (or the radius) of a cluster)
compute silhouette score for P_k
input: X : data
tree: ward tree
matric_measure ('euclidean', ...)
output: float array 1D size n
value of silhouette score of partion P_k
'''
n = len(X)
score = np.zeros(n-1)
print 'Length : ', n
for i in range(n-1):
#canot calculate the silhouette score for only one cluster
#should start from 2 clusters
k = i + 2
print '\n Cutting at k = ', k
label = _hc_cut(k,tree.children_, tree.n_leaves_)
print '\n Compute score ...'
s = metrics.silhouette_score(X, label, metric = metric_measure)
#s = silhouette_score_block(X, label, metric= metric_measure , sample_size=None )
score[k-2] = s
def compute_silhouette_block_tree(X, ward, metric_measure, verbose=False):
# extract all cut of H
num_cuts = ward.height_[len(ward.children_) + ward.n_leaves_ - 1]
if verbose == True:
print "Computing silhouette score :"
print "Height : ", num_cuts
sil = np.zeros(num_cuts)
for height_cut in np.arange(num_cuts):
if verbose:
print ".....scales: ", height_cut
num_clusters = len(ward.cut(height_cut))
label = _hc_cut(num_clusters, ward.children_, ward.n_leaves_)
iterations = 4
temp = np.zeros(interations)
for t in range(iterations):
score = silhouette_score_block(X, label, metric=metric_measure, sample_size=50000, n_jobs=2)
temp[t] = score
sil[height_cut] = temp[:].mean(0)
if verbose:
print ".....score: ", score
mean = sil[:].mean(0)
std = sil[:].std(0)
return mean, std
