np.zeros_like奇怪的bug

import numpy as np 

aa = np.array([[1,2,3],[2,3,3]])
cc = np.random.randn(2,3)
print(aa)
print(cc)

bb = np.zeros_like(aa)
print(bb)

for i in range(bb.shape[0]):
    for j in range(bb.shape[1]):
        bb[i,j] = cc[i,j]
        
print(bb)

结果如下

这里发现这个bb的结果是没有赋值的

正确做法如下

import numpy as np 

aa = np.array([[1,2,3],[2,3,3]])
cc = np.random.randn(2,3)
print(aa)
print(cc)

bb = np.zeros_like(aa,dtype=float)
print(bb)

for i in range(bb.shape[0]):
    for j in range(bb.shape[1]):
        bb[i,j] = cc[i,j]
        
print(bb)