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