09-10-2008 04:29 PM
Shortly after solving it I realized that it could be pared down to 160 solutions by rotating the star 60 degrees 5 times. From there, I guess you can pare it down to 80 solutions, because each solution has a mirror image that is technically the same star.
-D
09-10-2008 04:41 PM - edited 09-10-2008 04:43 PM
Its not a deterministic way to go but it can solve the puzzle quite fast.
09-10-2008 05:42 PM
ShotSimon wrote:Darren,
That is amazing and very well done. My guess is you are a puzzle master.
I've decided to leave Puzzle #3 open for you to choose...since I can't give away prizes
Try to keep the level of difficultly between LabVIEW Puzzle #1 and Puzzle #2. -SS
Thanks, I do enjoy puzzles, although I'd hardly consider myself a Puzzle Master (my unofficial score of 155 for the Google U.S. Puzzle Championship barely cracked the top 50). One that I tried to solve a while back and never got anywhere was the Round Robin problem posted on LAVA a few years ago. I kept getting stuck on the part mentioned by the original poster, where my code would hang when encountering a matchup that had already been generated by a previous iteration, even though the other matchups were different. Can anybody solve this one?
-D
09-11-2008 09:09 AM
BeCeGa,
Very nicely done as well.
I think if you combine Darrens way of doing the six comparisons (five are only needed if one row is pre-filled) with your solution you end up with a very simple solution. I also played around with mathscript commands such as randperm since when I originally solved this problem I was tormented for weeks and eventually solved it with MATLAB. -SS
09-11-2008 04:14 PM - edited 09-11-2008 04:18 PM
Darren,
OK not 100% yet but I may be close. In my example every two rows represents the round robin match-ups. I was thinking if you were to further randomize the matches, every two rows as a group. Then follows this by flipping the columns around, assuming these represent what field/court a person is on, then the result is a very random looking. If you flip Team 1 from the top to the bottom in the bottom half of the games before this randomization. You could use the top and the bottom to designate who wears what jersey, serves first, etc. Note when using an odd number of players the highest team number represents the by week for the opposing team.
Regards,
-SS
09-11-2008 04:26 PM
I think you're going to find that the randomization gets bogged down with large numbers of matchups, i.e. it's not as simple as switching around rows. Your algorithm for generating the initial matchups seems promising...let me know if you are able to enhance it to the point where it's generating random matchups for a given group number.
-D
09-12-2008 03:04 PM - edited 09-12-2008 03:05 PM
09-12-2008 03:29 PM - edited 09-12-2008 03:31 PM
LabVIEW Puzzle #4
Draw a 7x7 black and white board as shown and graphically solve for the longest closed uncrossed knight's tour for n = 7 (length 24). Assume you don't know the solution and are trying to find it. -SS
09-12-2008 03:39 PM
ShotSimon wrote:
Here is my new randomized version, let me know what you think -SS
I'm no mathematician, but that looks pretty random to me! 😛 Nice work, I'll post a link to this thread on LAVA in case the original poster is curious. As for your Knight's Path problem, don't expect a solution from me on that one anytime soon, I don't even know where to begin!
-D
09-12-2008 03:54 PM
Darren wrote:I'm no mathematician, but that looks pretty random to me! 😛 Nice work, I'll post a link to this thread on LAVA in case the original poster is curious. As for your Knight's Path problem, don't expect a solution from me on that one anytime soon, I don't even know where to begin!
Darren,
I think you could further enhance it by creating a fourth randomizer that would walk through each two row pair and randomly move the pairs around. I noticed that 1 will always be in the same column since I left out this step. -SS