A Julia package for computing clustering prototypes and criticisms.
In particular, we currently support the selection of prototypes and criticisms with the following methods:
Prototypes | Criticisms | |
---|---|---|
MMD-critic | ✓ | ✓ |
k-medoids | ✓ | ✓ |
k-means | ✓ | ✓ |
fuzzy c-means | ✓ | ✓ |
affinity propagation | ✓ | ✓ |
In order to compute prototypes and criticisms for clusterings, we provide two approaches:
- Directly use a clustering result computed with the Clustering.jl package.
- Provide the raw data set, the clustering assignments of the data instances, and the kind of method to be used for computing the prototypes and criticisms.
We assume that you have a working installation of Julia. This package is tested on macOS and Ubuntu (x64 platform, Julia 1.7 and nightly) by CI processes.
Open a Julia REPL and install the package:
julia>]
pkg>add https://github.com/laschuet/ClusteringPrototypesCriticisms.jl
Open a Julia REPL and start using the package:
julia>using ClusteringPrototypesCriticisms
Then you can easily read the API documentation:
julia>?
help?>prototypes
help?>criticisms
help?>sqmmd
help?>witness
Alternatively, we invite you to take a quick look at the source code.
We also provide some usage examples.