Page 6 of 6
Posted: Thu May 25, 2006 11:54 pm
by C
..my code has passed all the inputs above, but i still got WA. Is there any tricky cases ??... I feel helpless..
Posted: Sun May 28, 2006 8:13 pm
by Martin Macko
C wrote:..my code has passed all the inputs above, but i still got WA. Is there any tricky cases ??... I feel helpless..
I don't think there are any more tricky cases... If you're still getting WA, you can try to post your code here...
Posted: Sat Jun 03, 2006 9:20 pm
by ayon
Is this algorithm correct?
first we have n trees. now using dfs we find the line with the maximum number of trees(say p). if we have any other line with p trees we may choose anyone. fire those p trees. now we have n-p trees. with the same process now we find the best line with the maximum number of trees and fire as before until only m remaining.
is this process correct? if not please tell my why
Posted: Sat Jun 03, 2006 9:45 pm
by mf
No, it's wrong. Consider for example the trees, arranged as follows, you want to cut all of them:
Code: Select all
.012..
.3.4..
.567..
8.....
....9A
Optimal solution has 4 lines: 0-7-A, 2-5-8, 3-6-9, 1-4.
Your algorithm under some conditions could initially choose these lines: 0-3-5, 2-4-7; and then it'll need 3 more lines, for a total of 5 lines.
Posted: Sat Jun 03, 2006 9:56 pm
by ayon
oh, thank you very much for the wonderful example, i knew from previous posts that something like greedy wont work, but i could not find a good counter-example.
Re: 11008 - Antimatter Ray Clearcutting
Posted: Sat Jan 24, 2009 5:48 pm
by Sedefcho
With the risk of reviving this very old forum thread I want to proudly
mention
that I got my first ACC on this problem almost 3 years
after I first posted some pretty confused thoughts of mine (above).
I got back to this problem by pure chance and this time
(after doing some thinking in the last 1-2 days in order to
again understand the problem) I finally managed to
solve it (0.180 seconds in java)
I have used backtracking (instead of DFS), the rest
is the same as many people mentioned above
(simple geometry, bitmasks, memoization).
I don't think I have used any DP though.
This has been a pretty good problem, not saying
this just because I finally solved it, but because
of the many different techniques and considerations
involved in it (and in its solution).
Re: 11008 - Antimatter Ray Clearcutting
Posted: Tue Jan 03, 2012 9:53 am
by ankit1990
Can you please elaborate your approach slightly ?