You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some metrics take extra non-scalar arguments (e.g. vectors and arrays). There will need to be some adjustments both to the library and the tests to support this behavior. Have listed the metrics not yet implemented below.
With seuclidean and wminkowski, the functionality provided by these can be easily achieved by simply rescaling using the weights before passing it to euclidean or minkowski. So there is a fair question as to whether these are worth handling at all.
Have some WIP PRs for these three functions (mahalanobis, seuclidean, wminkowski) linked by the cross references above. These all need tests and may need some optimization.
Some metrics take extra non-scalar arguments (e.g. vectors and arrays). There will need to be some adjustments both to the library and the tests to support this behavior. Have listed the metrics not yet implemented below.
mahalanobis
( Add mahalanobis #58 )seuclidean
( Add seuclidean #56 )wminkowski
( Add wminkowski #57 )The text was updated successfully, but these errors were encountered: