TlMUR's blog

map VS unordered_map.
By TlMUR, history, 2 months ago, translation, In English

In yesterday's Div 3 contest in problem D, I used unordered_map to store elements, thinking it would be faster, but someone cracked my code, and my friend who had almost the same code, but used map instead of unordered_map, could not be cracked. Why was my code hacked, because unordered_map is faster than map, right? I have not been able to reach a clear answer to this question. So I ask from the respected Codeforces community, why did this happen? Thanks in advance for your answer.

  • Vote: I like it
  • +1
  • Vote: I do not like it

»
2 months ago, # |
  Vote: I like it 0 Vote: I do not like it

Auto comment: topic has been translated by timur_jalgasbaev (original revision, translated revision, compare)

→ Reply
»
2 months ago, # |
  Vote: I like it 0 Vote: I do not like it

unordered_map is basically O(1) for "random" inputs, but since the problemsetters/(people who make hacks) know that people will try using it, they can basically make an input that will "break" the unordered map and make it O(n) per access. Map on the other hand is always O(logn), and so its alot more safe to use.

→ Reply
»
2 months ago, # |
← Rev. 2   Vote: I like it +8 Vote: I do not like it

I did the exact same thing. This article explains very well why the unordered_map fails and also discusses the solution to avoid such hacks in the future.

→ Reply
»
2 months ago, # |
  Vote: I like it 0 Vote: I do not like it

Auto comment: topic has been updated by timur_jalgasbaev (previous revision, new revision, compare).

→ Reply
»
2 months ago, # |
← Rev. 4   Vote: I like it -19 Vote: I do not like it

Using gp_hash_table<int,int>mp works better than unordered_map because in the worst case, gp_hash_table operates in O(1), while unordered_map can degrade to O(n).Therefore, gp_hash_table is a better choise than unordered_map. If you use gp_hash_table, you need to include the following headers:

#include<ext/pb_ds/assoc_container.hpp>
#include <ext/pb_ds/tree_policy.hpp>
using namespace __gnu_pbds;
→ Reply
  • »
    »
    2 months ago, # ^ |
      Vote: I like it +8 Vote: I do not like it

    The article referenced above says it is even easier to hack than unordered_map.

    → Reply
  • »
    »
    2 months ago, # ^ |
    ← Rev. 2   Vote: I like it +8 Vote: I do not like it

    gp_hash_table can be hacked too.

    You can hack gp_hash_table by constructing a sequence $$$A$$$ that satisfies $$$A_i=65537i$$$.

    Then its time complexity will become $$$O(n)$$$.

    You can use this to prevent it from being hacked:

    struct myhash{
    static uint64_t rand(uint64_t x){
        x^=x<<13;
        x^=x>>7;
        return x^=x<<17;
    }
    size_t operator()(size_t x)const{
        static const uint64_t f_rand=chrono::steady_clock::now().time_since_epoch().count();
        return rand(x+f_rand);
    }
};
__gnu_pbds::gp_hash_table<size_t,size_t,myhash> hashtable;

    UPD: There is a typo in my code, so I fixed it.

    → Reply
  • »
    »
    2 months ago, # ^ |
    ← Rev. 2   Vote: I like it 0 Vote: I do not like it

    I see this for the first time, I haven’t even seen it in other people’s codes, I don’t think it’s optimal

    → Reply
»
2 months ago, # |
  Vote: I like it 0 Vote: I do not like it

Basically the unordered_map hack is that you put a lot of multiples of a special prime that'll make the unordered_map not being able to resize correctly. It'll put all of those numbers in a single "bucket" which is like a normal array, so searching in that array will take $$$O(n)$$$ time, thus causing the operation time to be $$$O(n)$$$. So your complexity goes up to $$$O(n^2)$$$ which is not what you want. To fix this problem you can use a custom hash function as described here, which will make the hash unpredictable so the hashes can't trigger the special prime.

→ Reply
»
2 months ago, # |
← Rev. 2   Vote: I like it 0 Vote: I do not like it

you could use mp.reserve(1024); mp.max_load_factor(1); which will make your unordered_map almost as fast as map and in some cases faster than map. you could try to resubmit the code using these two functions

→ Reply