scipy.spatial.distance.correlation

scipy.spatial.distance.correlation(u, v, w=None, centered=True)[source]

Computes the correlation distance between two 1-D arrays. The correlation distance between u and v, is defined as

\[1 - \frac{(u - \bar{u}) \cdot (v - \bar{v})} {{||(u - \bar{u})||}_2 {||(v - \bar{v})||}_2}\]

where \(\bar{u}\) is the mean of the elements of u and \(x \cdot y\) is the dot product of \(x\) and \(y\).

Parameters:

u : (N,) array_like

Input array.

v : (N,) array_like

Input array.

w : (N,) array_like, optional

The weights for each value in u and v. Default is None, which gives each value a weight of 1.0

Returns:

correlation : double

The correlation distance between 1-D array u and v.