Schrodinger_Bit's blog

By Schrodinger_Bit, 18 months ago, In English

Recently I was solving CSES-Towers . In this problem I firstly tried a solution using multiset O(nlogn) solution resulted in TLE. Then I optimized the solution with vector which also had a O(nlogn) complexity. I know solution with vector is the efficient one. But the other solution has also O(nlogn) complexity. Why it didn't work? Am I missing something ?

Solution with multiset :

#include<bits/stdc++.h>
using namespace std;
 
#define fastio ios_base::sync_with_stdio(false);cin.tie(NULL);cout.tie(NULL)
#define all(v) v.begin(), v.end()
 
int n,x;
multiset<int> s;
void solve(int cs){
    cin >> n;
    for(int i=0;i<n;i++){
        cin >> x;
        auto it=upper_bound(all(s),x);
        s.insert(x);
        if(it!=s.end()) s.erase(s.find(*it));
 
    }
    cout << s.size() << endl;
    
}    
signed main(){
    fastio;
    int T=1,cs=0;
    //cin >> T;
    while(T--) solve(++cs);
}

Solution with vector :

#include<bits/stdc++.h>
using namespace std;

#define pb push_back
#define all(v) v.begin(), v.end()
#define fastio ios_base::sync_with_stdio(false);cin.tie(NULL);cout.tie(NULL)

 
 
int n,x;
vector<int> s;
void solve(int cs){
    cin >> n;
    for(int i=0;i<n;i++){
        cin >> x;
        auto it=upper_bound(all(s),x);
        if(it==s.end()) s.pb(x);
        else s[it-s.begin()]=x;
 
    }
    cout << s.size() << endl;
    
}   
signed main(){
    fastio;
    int T=1,cs=0;
    //cin >> T;
    while(T--) solve(++cs);
}
  • Vote: I like it
  • +3
  • Vote: I do not like it

»
18 months ago, # |
  Vote: I like it +15 Vote: I do not like it

At here : auto it=upper_bound(all(s),x);

It is O(n),actually.

Use s.upper_bound(x) instead

  • »
    »
    18 months ago, # ^ |
      Vote: I like it +3 Vote: I do not like it

    Thanks a lot.

    It worked out. I didn't know about this.

  • »
    »
    12 months ago, # ^ |
      Vote: I like it 0 Vote: I do not like it

    Hi, I am curious to know how is auto it=upper_bound(all(s),x); O(n) here?

    from what I can see in the internal implementation, it uses __upper_bound() function which has a similar implementation to binary search:

    template<typename _ForwardIterator, typename _Tp, typename _Compare>
        _ForwardIterator
        __upper_bound(_ForwardIterator __first, _ForwardIterator __last,
    		  const _Tp& __val, _Compare __comp)
        {
          typedef typename iterator_traits<_ForwardIterator>::difference_type
    	_DistanceType;
    
          _DistanceType __len = std::distance(__first, __last);
    
          while (__len > 0)
    	{
    	  _DistanceType __half = __len >> 1;
    	  _ForwardIterator __middle = __first;
    	  std::advance(__middle, __half);
    	  if (__comp(__val, __middle))
    	    __len = __half;
    	  else
    	    {
    	      __first = __middle;
    	      ++__first;
    	      __len = __len - __half - 1;
    	    }
    	}
          return __first;
        }
    
    • »
      »
      »
      12 months ago, # ^ |
        Vote: I like it 0 Vote: I do not like it

      std::advance is linear for non-random access iterators.

      https://cplusplus.com/reference/iterator/advance/

      • »
        »
        »
        »
        12 months ago, # ^ |
          Vote: I like it 0 Vote: I do not like it

        It has always baffled me that the STL allows you to do it the wrong way, is there really no way they could've put some guardrails around this? Specializing the upper_bound for set iterators or something like that...

        The same thing for swapping sets, where you should do s1.swap(s2) instead of swap(s1, s2).

      • »
        »
        »
        »
        12 months ago, # ^ |
          Vote: I like it 0 Vote: I do not like it

        I get it now, thanks!