Python equivalent for C++ STL vector/list containers

Question

Is there something similar in Python that I would use for a container that's like a vector and a list?

Any links would be helpful too.

Gothurd

For linked list: [a0,[a1,[a2,[a3,[...]]]]] – Gerontocracy

Answer 1

You can use the inbuilt list - underlying implementation is similar to C++ vector. Although some things differ - for example, you can put objects of different type in one and the same list.

http://effbot.org/zone/python-list.htm

N.B.: Please keep in mind that vector and list are two very different data structures. List are heterogeneous, i.e. can store different object types, while C++ vectors are homogeneous. The data in vectors is stored in linear arrangement whereas in list is a collection of references to the type and the memory address of the variables.

Afterthought

Is accessing linear in time O(1), similar to accessing a vector as an array in C++? – Gothurd

Yes, underlying implementation is like a C++ vector. – Afterthought

That arrays are called lists (which is genereally used as shortcut for linked list, which is a wholly different data structure) is one of the few really unfortunate things in Python. – Seminarian

Thanks, it has the performance factors on the bottom, good info! – Gothurd

This answer is incomplete for being accepted: the list class is nothing like std::list. (There's no builtin linked list implementation in Python.) – Washbasin

Dead link. Could you provide another? – Indopacific

Answer 2

Have a look at Python's datastructures page. Here's a rough translation:

1. () => boost::Tuple (with one important distinction, you can't reassign values in a Python tuple)
2. [] => std::vector (as the comments have aluded towards, lacks memory characteristics associated with vectors)
3. [] => std::list
4. {} => tr1::unordered_map or boost::unordered_map (essentially a hash table)
5. set() => std::set

Needleful

Well, it can be... it depends on the backing implementation. But in CPython, yes, [] != std::list. – Much

@Amber: No Python implementation would ever dare to use linked lists for the builtin list type. That would totally screw about every piece of code that relies on indexing being O(1) (a totally valid assumption) - i.e. really much. We can safely ignore this scenario. – Seminarian

@delnan There are, however, backing data structures which would be much closer to a std::list than the CPython implementation. – Much

@Amber: Please elaborate. What data structure is "close to a linked list" but offers e.g. O(1) indexing and appending? – Seminarian

@delnan Certain variants of b-trees don't offer O(1) indexing, but can get fairly close (close enough that all but the largest edge cases wouldn't have a noticeable difference). – Much

@Amber: In reality it would be a big difference both in terms of performance and memory use, to the worse. So I don't see why anyone would want to differ from CPython in this case - after all, it's the de facto standard. – Afterthought

set() => std::unordered_set , since it's hashset – Hierarchize

Answer 3

py	cpp
deque	deque
PriorityQueue (or you may use heapq)	priorityqueue
set	unordered_set
list	vector
defaultdict(int)	unordered_map
list	stack
deque	queue
dict .get(val,0)	unordered_map
array	array
np.array	valarray

valarray is poor man's np.array.

in py >= 3.7, dict remember insert order. https://mcmap.net/q/75657/-how-do-you-retrieve-items-from-a-dictionary-in-the-order-that-they-39-re-inserted

---

In case you need TreeMap / TreeSet

https://github.com/grantjenks/python-sortedcontainers

Hierarchize

Answer 4

The cstl library wrapped commonly used C++ STL libraries including vector, unordered_map, and unordered_set for Python. It uses purely C++ implementation and does not have the copy-on-write issue that happens in all python objects. See https://github.com/fuzihaofzh/cstl for how to install and use.

Install it from pip

pip install cstl

Convert python objects into cstl objects:

import cstl

# Directly covert containers from python
v = cstl.frompy({"1":[1,2,3], "2":[4,5,6]}) # convert python object to cstl object
v["1"][2] = 10 # access cstl object
pv = cstl.topy(v) # convert cstl object to python object 
print(pv)

Vogel

Answer 5

Lists are sequences.

see http://docs.python.org/tutorial/datastructures.html

append is like push_back, see the other methods as well.

Pattiepattin

Answer 6

Python also has as part of the standard library an array type which is more efficient and the member type is constrained.

You may also look at numpy (not part of the standard library) if you need to get serious about efficient manipulation of large vectors/arrays.

Tret