Numpy Element Wise

---

layout: single title: Numpy - Element Wise Operations toc_label: Numpy - Element Wise Operations categories: [Python] tags: [Numpy, Broadcasting] author_profile: false search: true use_tex: true —

Element Wise Operations

Element Wise Operations

---

---

Scalar Operations

Scalar Addition

array + int array - int array * int array / int

for all element, add the specified value of integer

Array Operations

array1 + array2 array1 - array2 array1 * array2 array1 / arra2y

Dimension of both arrays are equal in the above array element-wise operations.

Broadcasting

Broadcasting operation is the mechanism that automatically adjusts the shapes of arrays with different size during arithmetic operations.

Example

A (2d array): 5 x 4
B (1d array): 1
Result (2d array): 5 x 4

A (2d array): 5 x 4
B (1d array): 4
Result (2d array): 5 x 4

A (3d array): 15 x 3 x 5
B (3d array): 15 x 1 x 5
Result (3d array): 15 x 3 x 5

A (3d array): 15 x 3 x 5
B (2d array): 3 x 5
Result (3d array): 15 x 3 x 5

A (3d array): 15 x 3 x 5
B (2d array): 3 x 1
Result (3d array): 15 x 3 x 5

a = np.array([[1, 2, 3], [4, 5, 6],[7, 8, 9]])
b = np.array([0, 1, 0])

print('Array "a":')
array_info(a)
print('Array "b":')
array_info(b)

print('Array "a+b":')
array_info(a + b)  # b is reshaped such that it can be added to a.

# b = [0,1,0] is broadcasted to     [[0, 1, 0],
#                                    [0, 1, 0],
#                                    [0, 1, 0]]  and added to a.

Array “a”:
[[1 2 3]
[4 5 6]
[7 8 9]]

Array “b”:
[0 1 0]

Array “a+b”:
[[1 3 3] [4 6 6] [7 9 9]]

Detach

detach()

• 역전파(gradient 계산)에서 해당 텐서를 분리합니다.

• 즉, 이 텐서는 계산 그래프에서 더 이상 연결되지 않음을 의미합니다.

• 주로, 학습 중에 모델의 출력값이나 중간 결과를 단순히 값으로만 쓰고 싶을 때 사용합니다.