SciPy

scipy.spatial.distance.cityblock

scipy.spatial.distance.cityblock(u, v, w=None)[source]

Compute the City Block (Manhattan) distance.

Computes the Manhattan distance between two 1-D arrays u and v, which is defined as

\[\sum_i {\left| u_i - v_i \right|}.\]
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:
cityblock : double

The City Block (Manhattan) distance between vectors u and v.

Examples

>>> from scipy.spatial import distance
>>> distance.cityblock([1, 0, 0], [0, 1, 0])
2
>>> distance.cityblock([1, 0, 0], [0, 2, 0])
3
>>> distance.cityblock([1, 0, 0], [1, 1, 0])
1

Previous topic

scipy.spatial.distance.chebyshev

Next topic

scipy.spatial.distance.correlation