Easy euclidean distance

[yupbank]

In [1]: import numpy as np
In [2]: from sklearn.metrics import pairwise_distances
In [3]: a = np.random.randn(10, 7)
In [4]: b = np.random.randn(5, 7)
In [5]: np.allclose(np.sqrt(np.linalg.norm(a, axis=1, keepdims=True)**2 - 2*np.matmul(a, b.T) + np.linalg.norm(b, axis=1, keepdims=True).T**2),
pairwise_distances(a, b))
Out[5]: True
In [6]:

[yupbank]

@tjhunter