VectorFeature processes high-dimensional vector data (like embeddings or feature
vectors) by applying k-means clustering to create discrete clusters that can be
monitored for distribution changes over time. This is particularly useful for
monitoring embedding drift in high-dimensional spaces.
The feature type is automatically set to CustomFeatureType.FROM_VECTOR and creates
meaningful groupings from vector data for drift detection and anomaly identification.
source_column
Optional original column if this feature is derived from an embedding
Examples
Creating a feature from a general embedding column:
Creating a feature from model hidden states:
Attributes
Type discriminator; fixed as CustomFeatureType.FROM_VECTOR.
Number of k-means clusters used for drift detection (default 5).
Cluster centroids in the embedded space; populated during training.
Name of the vector (embedding) column this feature is derived from.
classmethod validate_n_clusters()
Returns
int 