Why using erase and unique function to remove duplicate vectors from a 2d vector is adding an extra empty vector in 2d vector?

I am using the below code to remove duplicate vectors from a 2d vector

sort(final_vec.begin(), final_vec.end());
final_vec.erase(unique(final_vec.begin(), final_vec.end()));

Can someone explain me why this is happening and what change should I make.

>Solution :

To make it clearer what exactly is going wrong, I’m going to introduce an intermediate variable to store the iterator returned by std::unique.

Your code is equivalent to:

sort(final_vec.begin(), final_vec.end());
auto new_ending_iterator = unique(final_vec.begin(), final_vec.end())
final_vec.erase(new_ending_iterator);

When we check the overloads of std::vector::erase, we see that this is actually the one-argument erase call:

iterator erase( const_iterator pos )

1. Removes the element at pos.

So this is erasing a single element from final_vec. However, our goal is to erase every element from new_ending_iterator all the way until final_vec.end(). To do so, we use the second overload of erase:

iterator erase( const_iterator first, const_iterator last );

2. Removes the elements in the range [first, last).

The correct code would therefore look like

sort(final_vec.begin(), final_vec.end());
auto new_ending_iterator = unique(final_vec.begin(), final_vec.end())
final_vec.erase(new_ending_iterator, final_vec.end());

Or, if you really want to keep it all in one line:

sort(final_vec.begin(), final_vec.end());
final_vec.erase(unique(final_vec.begin(), final_vec.end()), final_vec.end());