CosineDistance<T>
Models & Types · AiDotNet.Clustering.DistanceMetrics
Computes Cosine distance between vectors.
Inheritance: DistanceMetricBase<T> → CosineDistance<T>
For Beginners
Cosine distance measures how different the “direction” of two vectors is, ignoring how long they are.
Example: Vectors [1, 0] and [0, 1] are perpendicular (90°), so their cosine similarity is 0 and distance is 1. Vectors [1, 2] and [2, 4] point in the same direction, so their distance is 0.
Best for:
- Text documents (TF-IDF vectors)
- Any data where magnitude doesn’t matter, only direction
- Sparse high-dimensional data
How It Works
Cosine distance measures the angle between two vectors, ignoring their magnitudes. It’s derived from cosine similarity: distance = 1 - similarity.
Formula: d(a, b) = 1 - (a · b) / (||a|| × ||b||) where a · b is the dot product and ||x|| is the L2 norm.
Constructors
| Constructor | Summary |
|---|---|
CosineDistance() |
Properties
| Property | Summary |
|---|---|
string Name { get; } |
Methods
| Method | Summary |
|---|---|
T Compute(Vector<T> a, Vector<T> b) |
|
T ComputeSimilarity(Vector<T> a, Vector<T> b) |
Computes cosine similarity (1 - distance). |