def cluster(X, fig, ax):
# Setup Affinity Propagation
af = AffinityPropagation(preference=-50).fit(X)
cluster_centers_indices = af.cluster_centers_indices_
labels = af.labels_
no_clusters = len(cluster_centers_indices)
print('Estimated number of clusters: %d' % no_clusters)
plot_columns = plotable(X)
centers = plotable(X[cluster_centers_indices])
plot1(no_clusters, labels, plot_columns, centers, ax)
return labels, no_clusters
def cluster(X, fig, ax):
num_clusters = int(input("Enter Number of cluster:"))
kmeans_model = KMeans(n_clusters=num_clusters, random_state=1, n_init=10)
kmeans_model.fit(X)
labels = kmeans_model.labels_
# clusters = {i: np.where(labels == i)[0] for i in range(kmeans_model.n_clusters)}
plot_columns = plotable(X)
centers = plotable(kmeans_model.cluster_centers_)
plot1(num_clusters, labels, plot_columns, centers, ax)
return labels, num_clusters
def cluster(X, fig, ax):
labels = ''
num_clusters = int(input("Enter Number of cluster:"))
knn_graph = kneighbors_graph(X, 30, include_self=False)
connectivity = (None, knn_graph)[0]
linkage = 'complete'
model = AgglomerativeClustering(linkage=linkage,
connectivity=connectivity,
n_clusters=num_clusters)
model.fit(X)
labels = model.labels_
plot_columns = plotable(X)
centers = None
plot1(num_clusters, labels, plot_columns, centers, ax)
return labels, num_clusters
def cluster(X, fig, ax):
num_clusters = int(input("Enter Number of cluster:"))
minibatch_model = MiniBatchKMeans(n_clusters=num_clusters,
init='k-means++',
n_init=1,
init_size=1000,
batch_size=1000)
minibatch_model.fit(X)
labels = minibatch_model.labels_
# clusters = {i: np.where(labels == i)[0] for i in range(minibatch_model.n_clusters)}
plot_columns = plotable(X)
centers = plotable(minibatch_model.cluster_centers_)
plot1(num_clusters, labels, plot_columns, centers, ax)
return labels, num_clusters
def cluster(X,fig,ax):
# Required for fuzzy cmeans(X is (num_feature,num_sample))
X = X.T
# print("Input shape:",X.shape)
num_cluster = int(input("Enter Number of cluster:"))
cntr, u_orig, _, _, _, _, _ = fuzz.cluster.cmeans(
X, num_cluster, 2, error=0.005, maxiter=1000)
X =X.T
plot_columns = plotable(X)
plot_centers = plotable(cntr)
# Labels for each sample
cluster_membership = np.argmax(u_orig, axis=0)
# Clusters
# clusters = {i: np.where(cluster_membership == i)[0] for i in range(num_cluster)}
plot1(num_cluster, cluster_membership, plot_columns, plot_centers, ax)
return cluster_membership, num_cluster