numpy-形状与形变

shape

data = np.array([[1, 2],
                 [3, 4]])

shape = data.shape
print(shape)        # (2, 2)

size

print(np.ones((3, 4)).size)    # 12

flatten：展开，拉成一维

data = np.array([[1, 2],
                 [3, 4]])

data1 = data.flatten()
print(data1)        # [1 2 3 4]

data2 = data.flatten('F')
print(data2)        # [1 3 2 4]

data3 = data.flatten('C')   ### 默认为 C
print(data3)        # [1 2 3 4]

reshape：形变

可以实现 flatten 效果

data = np.array([[1, 2],
                 [3, 4]])
data4 = np.reshape(data, (4, ))  ### 等价
print(data4)        # [1 2 3 4]

参考资料：