About dynamic allocation of vector pointers and vector pointers using shared_ptr

Dynamically allocated vector pointer

vector<int> *get_num(int n)
{
    vector<int> *pv = new vector<int>(n+1);
    int i , x;
    for(i = 0; i < n; i++)
    {
        cin>>x;
        <span style="background-color: rgb(255, 0, 0);">(*pv)[i] = (x);
</span>    }
    return pv;
}

If it is a dynamically allocated vector pointer, you can only use the array method (the above code is red) when inserting elements into the vector. If you use the .push_back () function method, you will get an error.

Vector pointer using shared_ptr

shared_ptr<vector<int>> get_num(int n)
{
    shared_ptr<vector<int>> pv = make_shared<vector<int>>();
    int i , x;
    for(i = 0; i < n; i++)
    {
        cin>>x;
        <span style="color:#ff0000;">(*pv).push_back(x);</span>
    }
    return pv;
}

If the vector pointer of shared_ptr is used, you can only use the .push_back method (the above code is red) when inserting elements into the vector. If you use the array insertion method, you will get an error. Know why?