def fit(self, X, y):
"""Compute supervised k-means clustering.
Parameters
----------
X : array-like of shape=(n_ts, sz, d)
Time series dataset.
y : array-like of shape=(n_ts,)
Time series labels to fit.
"""
cls, self.labels_ = np.unique(y, return_inverse=True)
self.n_clusters = len(cls)
if self.metric_params is None:
self.metric_params = {}
self.gamma_sdtw = self.metric_params.get("gamma_sdtw", 1.)
self.Xs_ = []
self.ys_ = []
centroids = []
for i in range(self.n_clusters):
self.Xs_.append(to_time_series_dataset(X[self.labels_ == i, :, :]))
self.ys_.append(self.labels_[self.labels_ == i])
if self.metric == 'euclidean':
centroids.append(EuclideanBarycenter().fit(self.Xs_[i]))
if self.metric == 'dtw':
centroids.append(DTWBarycenterAveraging().fit(self.Xs_[i]))
if self.metric == 'softdtw':
centroids.append(SoftDTWBarycenter().fit(self.Xs_[i]))
self.cluster_centers_ = np.stack([centroids]).squeeze()
return self
def _update_centroids(self, X):
for k in range(self.n_clusters):
if self.metric == "dtw":
self.cluster_centers_[k] = DTWBarycenterAveraging(max_iter=self.max_iter_barycenter,
barycenter_size=None,
init_barycenter=self.cluster_centers_[k],
verbose=False).fit(X[self.labels_ == k])
elif self.metric == "softdtw":
self.cluster_centers_[k] = SoftDTWBarycenter(max_iter=self.max_iter_barycenter,
gamma=self.gamma_sdtw,
init=self.cluster_centers_[k]).fit(X[self.labels_ == k])
else:
self.cluster_centers_[k] = EuclideanBarycenter().fit(X[self.labels_ == k])
def _update_centroids(self, X):
for k in range(self.n_clusters):
if self.metric == "dtw":
self.cluster_centers_[k] = dtw_barycenter_averaging(X=X[self.labels_ == k],
barycenter_size=None,
init_barycenter=self.cluster_centers_[k],
verbose=self.verbose)
# DTWBarycenterAveraging(max_iter=self.max_iter_barycenter,
# barycenter_size=None,
# init_barycenter=self.cluster_centers_[k],
# verbose=False).fit(X[self.labels_ == k])
elif self.metric == "dtwparallel":
self.cluster_centers_[k] = dtw_barycenter_averaging_parallel(X=X[self.labels_ == k],
barycenter_size=None,
init_barycenter=self.cluster_centers_[k],
verbose=self.verbose,
max_iter=10,
num_threads=self.num_threads)
elif self.metric == "softdtw":
self.cluster_centers_[k] = SoftDTWBarycenter(max_iter=self.max_iter_barycenter,
gamma=self.gamma_sdtw,
init=self.cluster_centers_[k]).fit(X[self.labels_ == k])
else:
self.cluster_centers_[k] = EuclideanBarycenter().fit(X[self.labels_ == k])
import numpy
import matplotlib.pyplot as plt
from tslearn.barycenters import EuclideanBarycenter, DTWBarycenterAveraging, SoftDTWBarycenter
from tslearn.datasets import CachedDatasets
numpy.random.seed(0)
X_train, y_train, X_test, y_test = CachedDatasets().load_dataset("Trace")
X = X_train[y_train == 2]
plt.figure()
plt.subplot(3, 1, 1)
for ts in X:
plt.plot(ts.ravel(), "k-", alpha=.2)
plt.plot(EuclideanBarycenter().fit(X).ravel(), "r-", linewidth=2)
plt.title("Euclidean barycenter")
plt.subplot(3, 1, 2)
dba = DTWBarycenterAveraging(max_iter=100, verbose=False)
dba_bar = dba.fit(X)
for ts in X:
plt.plot(ts.ravel(), "k-", alpha=.2)
plt.plot(dba_bar.ravel(), "r-", linewidth=2)
plt.title("DBA")
plt.subplot(3, 1, 3)
sdtw = SoftDTWBarycenter(gamma=1., max_iter=100)
sdtw_bar = sdtw.fit(X)
for ts in X:
plt.plot(ts.ravel(), "k-", alpha=.2)