numpy中的移位数组减法

Question

I have two arrays

A = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9])
B = np.array([6, 7, 8, 9, 1, 2, 3, 4, 5])

I need to subtract A from B but not in the normal way. I need to subtract 0th element of A from 4th element of B, 1st element of A from 5th element of B ie B[4] - A[0] , B[5] - A[1] , ... , B[n] - A[n-4] and so on. In short I need to shift elements of A by 4 indices and subtract from B and wrap the difference around. Is there a easy way to do this in python?

Answer 1

You can use numpy.roll :

numpy.roll(B, -4) - A

If you don't need to wrap around, you can use something like:

>>> B[4:] - A[:-4]
array([0, 0, 0, 0, 0])

Answer 2

如果将数组转换为Pandas Series，则可以使用shift（）方法。