Programming Fundamentals › Collections
Shallow vs Deep Copy
Copying only the top level vs copying everything nested inside.
Also known as: shallow copy, deep copy, cloning objects
Copying a value that contains other values has two levels:
- A shallow copy makes a new outer container, but the items inside are still the same objects as in the original.
- A deep copy copies everything nested inside too, so the two share nothing.
import copy
original = {"name": "Ana", "tags": ["a", "b"]}
shallow = copy.copy(original) # also: dict(original), original.copy()
deep = copy.deepcopy(original)
shallow["tags"].append("c")
print(original["tags"]) # ['a', 'b', 'c'] <- changed through the shallow copy!
print(deep["tags"]) # ['a', 'b']Changing shallow["name"] would be safe (strings are immutable), but the tags list is
shared. This is the classic surprise: you copied, edited the copy, and the original changed.
How to copy in each language
| Shallow | Deep | |
|---|---|---|
| Python | list(x), x[:], x.copy(), copy.copy(x) | copy.deepcopy(x) |
| JavaScript | { ...obj }, [...arr], Object.assign({}, obj) | structuredClone(obj) |
JSON.parse(JSON.stringify(obj)) is a common deep-copy trick, but it silently drops
undefined and functions and turns dates into strings. Prefer structuredClone in modern
runtimes (see structuredClone).
Which do you need?
Usually a shallow copy is enough, if you only replace values at the top level. If you change something nested, either deep copy first or copy each level you touch:
const next = { ...state, user: { ...state.user, name: "Bo" } };Deep copies cost time and memory on big structures, so don’t reach for them by reflex. Equal-looking copies are still different objects; see equality vs identity.