Removing vector elements?

[IVIax94]

Hey guys,

how can I remove (duplicate) vector elements using only .size(), .pop_back(), and .push_back, and without using a second vector?

an example of this (numbers between 1 and 10, length is 10 - this all depends on user input):

vector before: 10 7 10 4 1 0 10 4 9 7
vector after: 10 7 4 1 0 9

Any help would be appreciated! Thanks :)

[JLBorges]

> without using a second vector?

I suppose that means in O(1) size - ie. without using any auxiliary container at all.

If the order of the non-duplicated values need not be preserved, sort the vector, std::unique() (or its equivalent), and then pop off the duplicated values at the back of the sequence.
http://en.cppreference.com/w/cpp/algorithm/unique

If the order of the non-duplicated values must be preserved, something like this:

std::vector<int> seq = { 10, 7, 10, 4, 1, 0, 10, 4, 9, 7 } ;

for( std::size_t i = 0 ; i < seq.size() ; ++i )
{
    std::size_t j = i + 1 ;
    while( j < seq.size() ) // for every element after i
    {
        if( seq[i] == seq[j] ) // if it is a duplicate
        {
            // move all elements after it by one position to the left
            for( std::size_t k = j+1 ; k < seq.size() ; ++k )
               seq[k-1] = seq[k] ;
            seq.pop_back() ; // and pop off the last element
        }
        else ++j ;
    }
}