can, you need to increase precision of the accumulator in np.mean() or change the input dtype: >>> a.mean(dtype=np.float32) # default and lacks precision
確定！ 回上一頁
