OJ Board

this is a southwestern european regional competition (1995) problem
http://www.ntnu.edu.tw/acm/ProblemSetArchive/B_EU_SWERC/1995/

i have taken a look at the sample solution used to generate the outputs - it wrongly uses f as increasing y from bottom to top.

hence all the mistakes.

interestingly ...

did you remember to take out the freopen when you submitted?

besides, there are nC2 = n(n-1)/2 unique pairs for a set of n numbers. when you increment your total counter, you do not check whether i=j for this is not considered.

on a side note, you do not need to test input >input[j] for your gcd ...

did you remember to take out the freopen when you submitted?

besides, there are nC2 = n(n-1)/2 unique pairs for a set of n numbers. when you increment your total counter, you do not check whether i=j for this is not considered.

on a side note, you do not need to test input >input[j] for your gcd ...

Are you using pascal?
This input file is screwed for pascal.
I only managed to get it accepted in C.

Does your program give the correct output for the sample input?

Read the question carefully:

You are given the ranking of event i (ci) in chronological order, which means that
order[ci]=i; ci tells you the position of event i in chronological order. So your maximal increasing subsequence should ...

Why don't you try storing your own exponent and mantissa instead? How might you have used logs? I think you may have a floating point precision problem.

Positive integer means > 0

Use unsigned long. That is sufficient.

Do all calculations modulo 100000.