The following code confuses me. I know d has a different id with c, but why changing d lead to changes in c?
c=np.zeros(4)
d = c[:]
print(id(d))
print(id(c))
d[0] = 1
print(d)
print(c)
output:
2073635240496
2073589009424
[1. 0. 0. 0.]
[1. 0. 0. 0.]
The reason that happens is because of the layers of this object:
print(type(c))
print(type(c[0]))
Output:
<class 'numpy.ndarray'>
<class 'numpy.float64'>
c is a nested class.Let's look at the same situation with an easier to understand, nested list:
Without deepcopy:
a = [[3,8],[4,4]]
b = a[:]
print(a)
print(b)
b[0][0] = 9
print(a)
print(b)
Output:
[[3, 8], [4, 4]]
[[3, 8], [4, 4]]
[[9, 8], [4, 4]]
[[9, 8], [4, 4]]
With deepcopy:
from copy import deepcopy
a = [[3,8],[4,4]]
b = deepcopy(a)
print(a)
print(b)
b[0][0] = 9
print(a)
print(b)
Output:
[[3, 8], [4, 4]]
[[3, 8], [4, 4]]
[[3, 8], [4, 4]]
[[9, 8], [4, 4]]
When you set a variable to another variable that is a list, both variables are assigned to the same exact list, so changes in the list will appear in both variables. I would just replace d = c[:] to d = c[:].copy().
d = c[:] does a shallow copy so d = c[:].copy() is superfluous The issue arrises becuase of the way numpy builds it's array, and the OP needs to do a deepcopy of c in order to copy the nested arrays.