I found a bug in my code and while fixing that bug, and I noticed a peculiar behavior in numpy.
a = np.arange(10) # ints 0 -> 9
a[None] = 20
print(a) # output --> [20, 20, 20, 20, 20, 20, ... 20]
Why do the values in every position get replaced with 20?
This is the expected behavior, not a bug. a[None] is not slicing/indexing but broadcasting. What it does does is to provide a different view of the underlying data of a.
And since you are modifying this view with a[None]=20, you also modify a's data, and therefore a itself.