Re: Array vs Vector performance

JoshuaMaurice@[EMAIL PROTECTED]
 wrote:

> Even then, as Hak...@[EMAIL PROTECTED]
 pointed out, your test is comparing
> apples to oranges. std::vector stores its elements on the heap. Your
> array is statically allocated on the stack. The vector test is getting
> the heap memory management overhead, a.k.a. new[] and delete[] calls.
> If you change your array to be new[]'ed and delete[]'ed, you should
> see comparable results. A stack array will be much quicker than
> allocating an array on the heap, as your test demonstrates. If you
> know the size of the array at compile time, then use a stack array and
> not a vector. If you need the utility of vector (such as insert,
> erase, resize, etc.), and it's not performance critical, then use a
> vector.
> 
I think if you look at his code you will find that the
construction/destruction overheads have been extracted and that the
measurements are done on existing objects.

-- 
Note that robinton.demon.co.uk addresses are no longer valid.

      [ See http://www.gotw.ca/resources/clcm.htm
for info about ]
      [ comp.lang.c++.moderated.    First time posters: Do this! ]