cvi: Cluster Validity Indices in Python¶
These pages serve as the official documentation for the cvi Python package, the Python implementation of the ClusterValidityIndices.jl Julia package.
cvi provides batch and incremental cluster validity indices (CVIs) for evaluating
hard partitions when reference labels are unavailable.
The package offers a shared stateful interface for processing complete datasets (in a batch) or monitoring a clustering stream over time (incrementally).
Cluster Validity Indices (CVIs) tackle the problem of judging the performance of an unsupervised/clustering algorithm without the availability of truth or supervisory labels, resulting in metrics of under- or over-partitioning. Furthermore, Incremental CVIs (ICVIs) are variants of these ordinarily batch algorithms that enable an online and computationally tractable method of evaluating the performance of a clustering algorithm as it clusters while being numerically equivalent to their batch counterparts.
Start with Getting Started, then use Choosing an Index to select an index.
CONN has a specialized prototype-based interface covered in Using CONN.