943,910 Members | Top Members by Rank

Ad:
  • C++ Discussion Thread
  • Unsolved
  • Views: 2383
  • C++ RSS
You are currently viewing page 1 of this multi-page discussion thread
Oct 8th, 2008
0

number of ways to..ehh, i hate counting ways

Expand Post »
okay, this is the problem (variation of the knapsack coins problem): you've got 4 coin types:
1 cent, 2 cents, 5 cents and a quarter...infinite amount of each. I'm supposed to find the number of ways in which the coins can be arranged to form the sum of some integer n...
well i... had an idea but it proved pretty wrong.
my idea was to check all the possible ways of forming n with two other numbers, and save the solution as number of needed for the first times number of needed for the second, the problem is it is not always the case, as you can get large numbers cause of repetitions, for example 6:
2 + 4
two ways to make 2 ( 2, 1+1 )
three ways to make 4( 2+2, 1+1+1+1, 2+1 )
multiplied... 6 ways? nope...5
1+1+1+1+1+1
2+1+1+1+1
2+2+1+1
2+2+2
5+1... damn...
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

Fortunately, no need to count ways yourself. Let a dumb computer do that ...
Reputation Points: 1234
Solved Threads: 347
Postaholic
ArkM is offline Offline
2,001 posts
since Jul 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

hmm, yeah, but how can it be done in some reasonable time?
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

As usually: by solutions space reduction. Start from the largest coin type then recursively arrange the rest of sum with smaller set of coins and so on. Count all good solutions (with multiply and add operators)...
Reputation Points: 1234
Solved Threads: 347
Postaholic
ArkM is offline Offline
2,001 posts
since Jul 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

hmmm, could i please get a more detailed explanation?
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

i mean, for example 3 has 2 solutions: 1+1+1 and 2+1, so if i go your way, i'll first take 2 off of the once and run the recursion on 3-2=1
which will give 1 solution. then i'll go down to 1, and run recursion on 3-1=2, and2 will give me 2 solutions: 1+1, 2... so i end up with 3?
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 8th, 2008
0

Re: number of ways to..ehh, i hate counting ways

Nope. You have troubles with integer arithmetics.
For two cent coin: possible 0 or 1.
1st case: 3 - 0 = 3, for 1 cent coins 1 solution (1+1+1); 1 solution subtotal
2nd case: 3 - 2 = 1, for 1 cent coin 1 solution (1); 1 solution subtotal
1 + 1 = 2 - 2 solutions total, that's OK.
It is understandable that you hate counting ...
Last edited by ArkM; Oct 8th, 2008 at 6:37 pm.
Reputation Points: 1234
Solved Threads: 347
Postaholic
ArkM is offline Offline
2,001 posts
since Jul 2008
Oct 9th, 2008
0

Re: number of ways to..ehh, i hate counting ways

lol, i get it now finally... i thought i'm supposed to go from the largest coin down every time...that would be just sick.
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 9th, 2008
0

Re: number of ways to..ehh, i hate counting ways

hmm, well now i've implemented it, but it's so damn slow... it should work fast for 100k atleast... here's my implementation:
c++ Syntax (Toggle Plain Text)
  1. #include <cstdio>
  2. #include <cstring>
  3. #include <algorithm>
  4. #define MOD 100000
  5. using namespace std;
  6.  
  7. const int coins[] = {1, 2, 5, 25};
  8.  
  9. int n, dp[100000][5];
  10.  
  11. int rec(int n, int m) {
  12. int ret = 0; if(dp[n][m]) return dp[n][m];
  13. if(m == 0 || n == 0) return (dp[n][m] = 1);
  14. if(n < coins[m]) return (rec(n, m-1));
  15. for(int i = 0; i*coins[m] <= n; ++i) {
  16. ret = ((ret%MOD) + (rec(n-i*coins[m], m-1)%MOD))%MOD;
  17. }
  18. return (dp[n][m] = ret);
  19. }
  20.  
  21. int main( void )
  22. {
  23. scanf("%d", &n);
  24.  
  25. memset(dp, 0, sizeof dp);
  26. printf("%d\n", rec(n, 3)%MOD);
  27.  
  28. return 0;
  29. }
i guess my implementation is stupid in some way... or can it run no faster?
Last edited by gregorynoob; Oct 9th, 2008 at 3:54 pm.
Reputation Points: 12
Solved Threads: 5
Junior Poster in Training
gregorynoob is offline Offline
86 posts
since Jun 2008
Oct 9th, 2008
0

Re: number of ways to..ehh, i hate counting ways

Sorry, I have no time to inspect your implementation carefully, but no need in arrays in this algorithm (except coins which must be a parameter of a recursive function). No need in <algorithm> header or MOD macros too.
It must be a very fast (and short) code...
Last edited by ArkM; Oct 9th, 2008 at 8:55 pm.
Reputation Points: 1234
Solved Threads: 347
Postaholic
ArkM is offline Offline
2,001 posts
since Jul 2008

This thread is more than three months old

No one has posted to this discussion for at least three months. Please let old threads die and do not reply to them unless you feel you have something new and valuable to contribute that absolutely must be added to make the discussion complete. Otherwise, please start a new thread in this forum instead.
Message:





About Us | Contact Us | Advertise | Acceptable Use Policy
Forum Index | Build Custom RSS Feed


Follow us on Twitter


© 2011 DaniWeb® LLC