Time series distances#

The aeon.distances module contains time series specific distance functions that can be used in aeon and scikit learn estimators. It also contains tools for extracting the alignment paths for a distance calculation between two series, and tools for finding all pairwise distances

Dynamic Time Warping (DTW)#

dtw_distance(x, y[, window, itakura_max_slope])

Compute the DTW distance between two time series.

dtw_pairwise_distance(X[, y, window, ...])

Compute the DTW pairwise distance between a set of time series.

dtw_cost_matrix(x, y[, window, ...])

Compute the DTW cost matrix between two time series.

dtw_alignment_path(x, y[, window, ...])

Compute the DTW alignment path between two time series.

Move-Split-Merge (MSM)#

msm_distance(x, y[, window, independent, c, ...])

Compute the MSM distance between two time series.

msm_pairwise_distance(X[, y, window, ...])

Compute the msm pairwise distance between a set of time series.

msm_cost_matrix(x, y[, window, independent, ...])

Compute the MSM cost matrix between two time series.

msm_alignment_path(x, y[, window, ...])

Compute the msm alignment path between two time series.

Time Warp Edit (TWE)#

twe_distance(x, y[, window, nu, lmbda, ...])

Compute the TWE distance between two time series.

twe_pairwise_distance(X[, y, window, nu, ...])

Compute the TWE pairwise distance between a set of time series.

twe_cost_matrix(x, y[, window, nu, lmbda, ...])

Compute the TWE cost matrix between two time series.

twe_alignment_path(x, y[, window, nu, ...])

Compute the TWE alignment path between two time series.

Weighted Dynamic Time Warping (DTW)#

wdtw_distance(x, y[, window, g, ...])

Compute the WDTW distance between two time series.

wdtw_pairwise_distance(X[, y, window, g, ...])

Compute the WDTW pairwise distance between a set of time series.

wdtw_cost_matrix(x, y[, window, g, ...])

Compute the WDTW cost matrix between two time series.

wdtw_alignment_path(x, y[, window, g, ...])

Compute the WDTW alignment path between two time series.

Derivative Dynamic Time Warping (DDTW)#

ddtw_distance(x, y[, window, itakura_max_slope])

Compute the DDTW distance between two time series.

ddtw_pairwise_distance(X[, y, window, ...])

Compute the DDTW pairwise distance between a set of time series.

ddtw_cost_matrix(x, y[, window, ...])

Compute the DDTW cost matrix between two time series.

ddtw_alignment_path(x, y[, window, ...])

Compute the ddtw alignment path between two time series.

Weighted Derivative Dynamic Time Warping (WDDTW)#

wddtw_distance(x, y[, window, g, ...])

Compute the WDDTW distance between two time series.

wddtw_pairwise_distance(X[, y, window, g, ...])

Compute the WDDTW pairwise distance between a set of time series.

wddtw_cost_matrix(x, y[, window, g, ...])

Compute the WDDTW cost matrix between two time series.

wddtw_alignment_path(x, y[, window, g, ...])

Compute the WDDTW alignment path between two time series.

Longest Common Subsequence (LCSS)#

lcss_distance(x, y[, window, epsilon, ...])

Return the LCSS distance between x and y.

lcss_pairwise_distance(X[, y, window, ...])

Compute the LCSS pairwise distance between a set of time series.

lcss_cost_matrix(x, y[, window, epsilon, ...])

Return the LCSS cost matrix between x and y.

lcss_alignment_path(x, y[, window, epsilon, ...])

Compute the LCSS alignment path between two time series.

Edit Real Penalty (ERP)#

erp_distance(x, y[, window, g, g_arr, ...])

Compute the ERP distance between two time series.

erp_pairwise_distance(X[, y, window, g, ...])

Compute the ERP pairwise distance between a set of time series.

erp_cost_matrix(x, y[, window, g, g_arr, ...])

Compute the ERP cost matrix between two time series.

erp_alignment_path(x, y[, window, g, g_arr, ...])

Compute the ERP alignment path between two time series.

Edit distance for real sequences (EDR)#

edr_distance(x, y[, window, epsilon, ...])

Compute the EDR distance between two time series.

edr_pairwise_distance(X[, y, window, ...])

Compute the pairwise EDR distance between a set of time series.

edr_cost_matrix(x, y[, window, epsilon, ...])

Compute the EDR cost matrix between two time series.

edr_alignment_path(x, y[, window, epsilon, ...])

Compute the EDR alignment path between two time series.

General methods with distance argument#

distance(x, y, metric, **kwargs)

Compute the distance between two time series.

pairwise_distance(x[, y, metric])

Compute the pairwise distance matrix between two time series.

cost_matrix(x, y, metric, **kwargs)

Compute the alignment path and distance between two time series.

alignment_path(x, y, metric, **kwargs)

Compute the alignment path and distance between two time series.

General methods to recover distance functions#

get_distance_function(metric)

Get the distance function for a given metric string or callable.

get_pairwise_distance_function(metric)

Get the pairwise distance function for a given metric string or callable.

get_cost_matrix_function(metric)

Get the cost matrix function for a given metric string or callable.

get_alignment_path_function(metric)

Get the alignment path function for a given metric string or callable.