OJ Board

Some Correct input and output :

Input :

10 20
23 62 15 28 72 83 56 80 30 30 98 82 51 76 39 21 65 83 72 9
1 2 1 3 2 4 2 5 4 6 4 7 6 8 3 9 7 10
7 5
4 9
1 2
10 2
9 8
6 9
6 5
9 7
1 4
8 2
8 6

10 20
41 40 10 85 9 57 37 43 17 92 55 32 47 21 69 44 87 23 8 78
1 2 2 3 3 4 1 5 3 6 5 7 4 8 3 9 4 10
8 1 ...

Hi I tried to solve this problem but I can not proceed any more, can you give me a hint?
I was thinking like this : dp[node][t] means from node and it's subtree the t-th largest value which also satisfies the conditions.
then I can update dp [k] using it's children. O( n * k * log(k) )
But I can not ...

By the way, Thank you, I changed it a little bit and got ACCEPTED

That's so strange, last night I was testing my program with Visual Studio 2010, and it worked very good, generated correct output as I said in my previous post,
but now, after your reply I tested it with MinGW/G++ and it didn't work, I checked my code again and noticed line 19 :
return dp[v] = f ...

Yeah Thank you, I got TLE exceeded too (using hash), I solved it using TRIE.
Do you have any idea on the dp approach of mak(cse_DU)?

Has anybody solved this problem using Hashing??? I don't think the running time of hashing is good enough to pass all tests.
I know other ways of solving this problem, I'm just curious about solving it using hashing.
They have tagged it as "hashing" in this page http://uvatoolkit.com/problemssolve ...

Just check for bases up to 50, and the only special case is when base = 1 (which is clearly explained in the problem statement).

I checked it for small n, then I found a pattern, but I have no proof for it.

Input : Output :

neither n nor m is zero in the judge data.

I have a problem here :
When I produce all possible inputs and I put it in my source code just to print in UVa OJ it gets ACCEPTED.

....
mat[25][4] = 114079985111LL;
mat[26][1] = 1LL;
mat[26][2] = 196418LL;
mat[26][3] = 21489003LL;
....

But when I ...