The difference between vector and deque [duplicate]

Answer 1

Performance, mainly. An std::deque has all of the functionality of std::vector, at least for normal use, but indexing and iterating over it will typically be somewhat slower; the same may hold for appending at the end, if you've used reserve. And of course, std::vector is the default container, using anything else will suggest to the reader that you have special requirements.

std::vector also guarantees contiguity, so it (and only it) can be used to interface legacy functions which require a T* or a T const*.

I might add that the one time I actually had a performance issue, and measured, std::vector was faster than std::deque, despite the fact that I was regularly removing elements from the front (using the container as a queue, pushing at the back, and popping at the front). I don't know if that generalizes, however; in my case, the queue was relatively short (never more than about 15 elements, and usually many less), and the contents were char, which is extremely cheap to copy. But in general, I'd use std::vector even if I needed to remove elements from the front, if only because of its better locality. I'd probably only consider std::deque if I expected thousands of elements, which were expensive to copy.

Answer 2

std::deque is a double-ended queue. It provides efficient insertion and deletion of elements at the beginning also, not just at the end, as std::vector does. Vectors are guaranteed to store the elements in a contiguous storage, therefore you can access its elements by index/offset. std::deque does not offer this guarantee.

Answer 3

One main difference between vectors and deques is that the latter allows efficient insertion at the front of the structure as well as the back.

Deques also do not guarantee that their elements are contiguous in memory so the at-style operator (indexing) may not be as efficient.

Note that the difference is unlikely to matter in practice for smaller collections but would generally become more important if, for example, the collection size increases or you're modifying it many times per second.