jmrh_1's blog

By jmrh_1, history, 5 years ago, In English

Hello codeforces Let's say an F(string s) as the number of different substrings that s has. Which is the maximum F(string x) such that x at most has 100000 characters and a dictionary of 26 characters.

  • Vote: I like it
  • +32
  • Vote: I do not like it

| Write comment?
»
5 years ago, # |
  Vote: I like it 0 Vote: I do not like it

Is there a problem link

»
5 years ago, # |
  Vote: I like it -20 Vote: I do not like it

https://cp-algorithms.com/string/string-hashing.html in this tutorial you can learn how to find the number of distinct substrings using hashing.

»
5 years ago, # |
  Vote: I like it -14 Vote: I do not like it

To solve this you can build a suffix tree of the original string S and count the amount of nodes on the tree. The complexity of this solution is O(N) if you use Ukkonen's algorithm to build the suffix tree.

»
5 years ago, # |
  Vote: I like it -15 Vote: I do not like it

here is the problem link

and here is the solution by me.

»
5 years ago, # |
Rev. 2   Vote: I like it +16 Vote: I do not like it

I know that there is a solution in O (N log N) with Suffix Array and LCP, but the real question is the maximum value that the F () function could reach in any string of at most 100,000 characters with a 26 letter dictionary.

»
5 years ago, # |
  Vote: I like it +5 Vote: I do not like it

I think that the more characters the string has, the less possibility there is that the function F () defined for every string of at most 100000 and with a 26-letter dictionary reaches its maximum value but I am not sure of the maximum value of the function F (). Which would suit me very well because I think I solve many problems depending on the maximum value of F ().

»
5 years ago, # |
Rev. 2   Vote: I like it +6 Vote: I do not like it

I wrote the Suffix Array solution, and generated a few random strings of length 100.000.

The maximal value of the number of substrings is N * (N - 1) / 2 = 5000050000, and all the random strings i generated had around 4999700000 distinct substrings. It's almost 0.99995% of the maximal value we could get.

If we think about it it's quite obvious why: as the string is random, there is virtually no chance that two different substrings of length >= 10 match (the probability of a colision is 1/26^10 = 7e-15 ~= 0). So except for the small substrings of length 1, 2, ... 9 (which are 9 * N = 900000), all the remaining 4999500000 are distinct.

So if your hope was that there are few such distinct substrings, sorry to break it down :(

  • »
    »
    5 years ago, # ^ |
      Vote: I like it +14 Vote: I do not like it

    Moreover, there's an obvious statement that there're no more than $$$26$$$ distinct substrings of size 1, $$$26^2$$$ of size 2 and $$$26^3$$$ of size 3. N * (N - 1) // 2 - (N - 26) - (N - 1 - 26 * 26) - (N - 2 - 26 * 26 * 26) equals 4999668281. You get almost that (if not exactly).