Why does NumPy array not update original array using predictor-CodePudding

I am learning NumPy (and Python) and working through some exercises regarding arrays. I had a question that came up that I could not comprehend. I understand that the following code should update the original array that b is pointed to. The code is as follows.

a = np.zeros((4,5))
b = a[1:3,0:3]
b = b   100
print('a = ', a)
print('b = ', b)

The output is:

a =  [[0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0.]
 [0. 0. 0. 0. 0.]]
b =  [[100. 100. 100.]
 [100. 100. 100.]]

Why does a not update along with b? I understand that a pointer object should update the original with edits. I'm assuming it is due to the syntax that 100 is added to b. The code below updates, as I thought, with the original array changing. What is the difference?

a = np.zeros((4,5))
b = a[1:3,0:3]
b[0, 0] = 100
print('a = ', a)

Output:

a =  [[  0.   0.   0.   0.   0.]
 [100.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.]]

CodePudding user response：

Python does not have pointers, at least not in the sense of other languages such as C.

What is happening here is that when you index an array, i.e. a[1:3,0:3], you create a view of a part of the data in a. Numpy docs on what views are

In your first example, you then evaluate b 100, which creates a new array. This behavior is not changed by the fact that you assign the result to the same variable name b.

In your second example b[0, 0] = 100, you modify b in-place. Since b is just a view of a, you also modify corresponding entries of a.

You could rewrite your first example to do b = 100 instead, which again modifies b in-place and will also change a

CodePudding user response：

In the first code block, you are setting the value of b equal to what it was before and adding 100 to all of its value.

b = b   100

In the second code block, you are setting the value of the number in the [0, 0] position equal to 100. Since b is just a pointer to a, it also modifies the value of a

b[0, 0] = 100

The difference between the two is that in the first one you are cloning b then modify it (then assign it to b). Where in the second one, you do not clone it, and just modifies it.