get_pairwise_distance_function

get_pairwise_distance_function(metric: str | Callable[[ndarray, ndarray, Any], ndarray]) → Callable[[ndarray, ndarray, Any], ndarray]

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

metric	Distance Function
‘dtw’	distances.dtw_pairwise_distance
‘shape_dtw’	distances.shape_dtw_pairwise_distance
‘ddtw’	distances.ddtw_pairwise_distance
‘wdtw’	distances.wdtw_pairwise_distance
‘wddtw’	distances.wddtw_pairwise_distance
‘adtw’	distances.adtw_pairwise_distance
‘erp’	distances.erp_pairwise_distance
‘edr’	distances.edr_pairwise_distance
‘msm’	distances.msm_pairiwse_distance
‘twe’	distances.twe_pairwise_distance
‘lcss’	distances.lcss_pairwise_distance
‘euclidean’	distances.euclidean_pairwise_distance
‘squared’	distances.squared_pairwise_distance
‘manhattan’	distances.manhattan_pairwise_distance
‘minkowski’	distances.minkowski_pairwise_distance
‘sbd’	distances.sbd_pairwise_distance

Parameters:

metricstr or Callable

The metric string to resolve to a alignment path function. If string given then it will be resolved to a alignment path function. If a callable is given, the value must be a function that accepts two numpy arrays and **kwargs returns a np.ndarray that is the pairwise distance between each time series.

Returns:

Callable[[np.ndarray, np.ndarray, Any], np.ndarray]

The pairwise distance function for the given metric.

Raises:

ValueError

If metric is not one of the supported strings or a callable.

Examples

>>> from aeon.distances import get_pairwise_distance_function
>>> import numpy as np
>>> x = np.array([[[1, 2, 3]],[[4, 5, 6]], [[7, 8, 9]]])
>>> y = np.array([[[11, 12, 13]],[[14, 15, 16]], [[17, 18, 19]]])
>>> dtw_pairwise_dist_func = get_pairwise_distance_function("dtw")
>>> dtw_pairwise_dist_func(x, y, window=0.2)
array([[300., 507., 768.],
       [147., 300., 507.],
       [ 48., 147., 300.]])