numpy的rank1 array与矩阵内积外积计算

a = np.random.randn(5)
print(a)
## [-1.17124494 -1.0144042  -1.81090015  0.06239375  0.80871696]
print(a.T)
## [ 1.27831203 -0.02878799 -0.18697777  0.22739936 -0.53940577]
## a为rank1 vector, 无法进行outer product计算

b = np.random.randn(5, 1)
print(b) # b为column vector
# [[-0.76206967]
#  [ 0.38957009]
#  [-0.60387659]
#  [ 1.32058981]
#  [ 0.28547156]]
c = np.random.randn(1, 5)
print(c) # c为row vector
# [[-0.12683617 -0.07037879  0.68081286 -0.93557581  1.1530988 ]]
dot = np.dot(b,c)
outer = np.outer(b,c)
print(dot.shape) 
# (5, 5)
print(outer.shape) 
# (5, 5)
# 可见进行outer product运算的结果相同