TimeSeriesKMeans

Time series K-mean implementation.

Quickstart

from sktime.clustering.k_means import TimeSeriesKMeans

estimator = TimeSeriesKMeans(n_clusters: int=8, init_algorithm: str | Callable='random', metric: str | Callable='dtw', n_init: int=10, max_iter: int=300, tol: float=1e-06, verbose: bool=False, random_state: int | RandomState=None, averaging_method: str | Callable [[ndarray ], ndarray ]='mean', distance_params: dict=None, average_params: dict=None)

Parameters(11)

n_clusters: int, defaults = 8

The number of clusters to form as well as the number of centroids to generate.

init_algorithm: str, np.ndarray (3d array of shape (n_clusters, n_dimensions,

series_length)), defaults = ‘random’ Method for initializing cluster centers or an array of initial cluster centers. If string, any of the following strings are valid:

[‘kmeans++’, ‘random’, ‘forgy’].

If 3D np.ndarray, initializes cluster centers with the provided array. The array

must have shape (n_clusters, n_dimensions, series_length) and the number of clusters in the array must be the same as what is provided to the n_clusters argument.

metric: str or Callable, defaults = ‘dtw’

Distance metric to compute similarity between time series. Any of the following are valid: [‘dtw’, ‘euclidean’, ‘erp’, ‘edr’, ‘lcss’, ‘squared’, ‘ddtw’, ‘wdtw’, ‘wddtw’]

n_init: int, defaults = 10

Number of times the k-means algorithm will be run with different centroid seeds. The final result will be the best output of n_init consecutive runs in terms of inertia.

max_iter: int, defaults = 300

Maximum number of iterations of the k-means algorithm for a single run.

tol: float, defaults = 1e-6

Relative tolerance with regards to Frobenius norm of the difference in the cluster centers of two consecutive iterations to declare convergence.

verbose: bool, defaults = False

Verbosity mode.

random_state: int or np.random.RandomState instance or None, defaults = None

Determines random number generation for centroid initialization.

averaging_method: str or Callable, defaults = ‘mean’

Averaging method to compute the average of a cluster. Any of the following strings are valid: [‘mean’, ‘dba’]. If a Callable is provided must take the form Callable[[np.ndarray], np.ndarray].

average_params: dict, defaults = None = no parameters

Dictionary containing kwargs for averaging_method.

distance_params: dict, defaults = None = no parameters

Dictionary containing kwargs for the distance metric being used.

Examples

>>> from sktime.datasets import load_arrow_head
>>> from sktime.clustering.k_means import TimeSeriesKMeans
>>> X_train, y_train = load_arrow_head (split = "train")
>>> X_test, y_test = load_arrow_head (split = "test")
>>> clusterer = TimeSeriesKMeans (n_clusters = 3)
>>> clusterer. fit (X_train) TimeSeriesKMeans(n_clusters=3)
>>> y_pred = clusterer. predict (X_test)