The traditional implementation of K-Means algorithm is usually considering only the euclidean distance to compute the distance between the points in the feature space. This new libary is an implementation of the K-Means algorithm with multiple distances choices:

• Euclidean Distance
• Cosine Distance
• City-Block Distance
• L1 Distance
• L2 Distance
• Manhattan Distance
• Bray-Curtis Distance
• Canberra Distance
• Chebysev Distance
• Correlation Distance
• Mahalanobis Distance
• Seuclidean Distance
• Sqeuclidean Distance

In order to install the library, use pip:

pip install kmeans_multidistance

To see an introduction to using the library, please check the demo notebook Demo Notebook