224

What is the cheapest way to initialize a std::vector from a C-style array?

Example: In the following class, I have a vector, but due to outside restrictions, the data will be passed in as C-style array:

class Foo {
  std::vector<double> w_;
public:
  void set_data(double* w, int len){
   // how to cheaply initialize the std::vector?
}

Obviously, I can call w_.resize() and then loop over the elements, or call std::copy(). Are there any better methods?

MD XF
  • 7,860
  • 7
  • 40
  • 71
Frank
  • 64,140
  • 93
  • 237
  • 324
  • 12
    The crux of the problem is that there is no way for the vector to know if the same allocator was used to create your C-style array. As such the vector must allocate memory using its own allocator. Otherwise it could simply swap out the underlying array and replace it with your array. – Void - Othman Mar 12 '10 at 18:04

6 Answers6

301

Don't forget that you can treat pointers as iterators:

w_.assign(w, w + len);
Pavel Minaev
  • 99,783
  • 25
  • 219
  • 289
  • Is there a performance difference between `std::copy(w, w + len, w_.begin())` and your `assign` solution? – Frank Mar 12 '10 at 19:37
  • 1
    Oh, I guess the difference is that `std::copy` will not resize the array. – Frank Mar 12 '10 at 20:01
  • I don't know if *assign* is smart enough to compute how much *w* is distant from *w+len* (a generic iterator AFAIK does not provide a quick way to do this), so you may enhance a little the performances putting a w_.reserve(len) before that statement; in the worst case you gain nothing. In this way it should have more or less the performance of resize+copy. – Matteo Italia Mar 12 '10 at 21:25
  • 3
    It's a quality of implementation issue. Since iterators have tags that specify their categories, an implementation of `assign` is certainly free to use them to optimize; at least in VC++, it does indeed do just that. – Pavel Minaev Mar 14 '10 at 01:33
  • 50
    The quick solution could be std::vector w_(w,w+len); – jamk May 15 '13 at 12:13
  • 7
    This copies elements to a newly created storage for 'w_'; 'w_.data' will not point to 'w'. You still have to deallocate 'w'. There's no ownership transfer – Indy9000 May 30 '14 at 14:03
  • Sure, but this is as good as you can get with a vector. It does not provide any facility to attach to an existing buffer. – Pavel Minaev May 31 '14 at 20:48
  • Is this safe to do when when `w+len` points to the element past the end of the array, as in, when you want to use this method to take the last 10 elements of a 100 element array? In VS2013 I got a debug assertion that the source array had been accessed out of bounds, which made me cautious, and I ended up using a more C-style approach. – Wayne Uroda May 20 '15 at 00:20
  • 1
    If it's _one_ element past the end, it should be okay (just as `v.end()` is an iterator pointing one past the end with vector in a similar case). If you do get an assertion, then something is off elsewhere. – Pavel Minaev May 20 '15 at 04:50
  • Also if you want to statically allocate in C++11 use initializer_list like below vector w_{1.2, 2.3, 4.3}; – Vineel Jul 02 '16 at 06:22
  • 3
    Just quickly, will this deallocate the array memory when the vector goes out of scope? – Adrian Albert Koch Nov 15 '19 at 17:23
49

You use the word initialize so it's unclear if this is one-time assignment or can happen multiple times.

If you just need a one time initialization, you can put it in the constructor and use the two iterator vector constructor:

Foo::Foo(double* w, int len) : w_(w, w + len) { }

Otherwise use assign as previously suggested:

void set_data(double* w, int len)
{
    w_.assign(w, w + len);
}
Mark B
  • 95,107
  • 10
  • 109
  • 188
18

The quick generic answer:

std::vector<double> vec(carray,carray+carray_size);

or question specific: 

std::vector<double> w_(w,w+len);

based on above: Don't forget that you can treat pointers as iterators

imbr
  • 6,226
  • 4
  • 53
  • 65
17

Well, Pavel was close, but there's even a more simple and elegant solution to initialize a sequential container from a c style array.

In your case:

w_ (array, std::end(array))
  • array will get us a pointer to the beginning of the array (didn't catch it's name),
  • std::end(array) will get us an iterator to the end of the array.
Mugurel
  • 1,829
  • 3
  • 17
  • 26
  • 1
    What includes/version of C++ does this require? – Vlad Nov 12 '15 at 16:18
  • 1
    This is one of the constructors of std::vector from at least c++98 onwards.... It's called 'range constructor'. http://www.cplusplus.com/reference/vector/vector/vector/ Try it. – Mugurel Nov 12 '15 at 17:49
  • 2
    More independent version is: w_ (std::begin(array), std::end(array)); (In the future you can to change a C array for a C++ container). – Andrew Romanov Feb 15 '16 at 08:20
  • 18
    Mind you, this only works if you have a real `array` (which usually means you're copying from a global or local (declared in current function) array). In the OP's case, he's receiving a pointer and a length, and because it's not templated on the length, they can't change to receiving a pointer to a sized array or anything, so `std::end` won't work. – ShadowRanger Jun 23 '16 at 23:12
  • 2
    `vector` does not overload `operator()`, so this won't compile. `std::end` being called on a pointer is no use either (the question asks to assign a vector from a pointer and a separate length variable). It would improve your answer to show more context about what you are trying to suggest – M.M Mar 09 '17 at 19:41
12

You can 'learn' the size of the array automatically:

template<typename T, size_t N>
void set_data(const T (&w)[N]){
    w_.assign(w, w+N);
}

Hopefully, you can change the interface to set_data as above. It still accepts a C-style array as its first argument. It just happens to take it by reference.

How it works

[ Update: See here for a more comprehensive discussion on learning the size ]

Here is a more general solution:

template<typename T, size_t N>
void copy_from_array(vector<T> &target_vector, const T (&source_array)[N]) {
    target_vector.assign(source_array, source_array+N);
}

This works because the array is being passed as a reference-to-an-array. In C/C++, you cannot pass an array as a function, instead it will decay to a pointer and you lose the size. But in C++, you can pass a reference to the array.

Passing an array by reference requires the types to match up exactly. The size of an array is part of its type. This means we can use the template parameter N to learn the size for us.

It might be even simpler to have this function which returns a vector. With appropriate compiler optimizations in effect, this should be faster than it looks.

template<typename T, size_t N>
vector<T> convert_array_to_vector(const T (&source_array)[N]) {
    return vector<T>(source_array, source_array+N);
}
Community
  • 1
  • 1
Aaron McDaid
  • 26,501
  • 9
  • 66
  • 88
  • 1
    In the last sample, `return { begin(source_array), end(source_array) };` is also possible – M.M Mar 09 '17 at 19:45
0

std::vector<double>::assign is the way to go, because it's little code. But how does it work, actually? Doesnt't it resize and then copy? In MS implementation of STL I am using it does exactly so.

I'm afraid there's no faster way to implement (re)initializing your std::vector.

Janusz Lenar
  • 1,690
  • 2
  • 13
  • 19