View previous topic :: View next topic |
Author |
Message |
| mrmarky2
| Joined: 13 Jan 2007 | Posts: 64 | : | | Items |
|
Posted: Tue May 13, 2008 3:14 pm Post subject: Colouring |
|
|
http://www.scanraid.com/Singles_Chains
http://www.sudocue.net/guide.php#ConnectedPairs
Im having a few problems working out how to put colouring into my program. The links above are where I found colouring, or singles chains.
Anyone have an idea on how to look for it?
I have not been programming long, so I dont understand a lot of things. Could someone try and explain it to me simply?
Thanks
Mark |
|
Back to top |
|
|
| mrmarky2
| Joined: 13 Jan 2007 | Posts: 64 | : | | Items |
|
Posted: Tue May 13, 2008 3:23 pm Post subject: |
|
|
I think I might have an idea how to do it:
1. Look for candidates that are the last 2 in each house.
2. Use them to find clusters.
3. Look at the clusters greater than 4.
4. Look at the two parity situations.
5a. Eliminate pencilmarks of that number that are in the same house as 2 different parity cases.
5b. Eliminate pencilmarks of that number that can see two parities.
5c. See if a cluster destroys itself.
Is this the best way to go about it?
Thanks
Mark |
|
Back to top |
|
|
| Lunatic
| Joined: 11 Mar 2007 | Posts: 166 | : | Location: Ghent - Belgium | Items |
|
Posted: Sat May 17, 2008 2:24 pm Post subject: |
|
|
mrmarky wrote: | 3. Look at the clusters greater than 4. |
I think you should look at clusters greater than 3
Cell1---->Cell2---->Cell3---->Cell4
Any cell seen by both Cell1 and Cell4 (or Cell6, Cell8, Cell10, and so on...) cannot longer have the candidate under examination.
But...
Cell1---->Cell2---->Cell3---->Cell4---->Cell5
If Cell1 and Cell5 (or Cell7, Cell9, Cell11, and so on...) are seeing each other then all 'odd' labeled cells are false and loose the candidate under examination. _________________ Marc
~~~<><~~~<><~~~<><~~~<><~~~ |
|
Back to top |
|
|
| mrmarky2
| Joined: 13 Jan 2007 | Posts: 64 | : | | Items |
|
Posted: Sun May 18, 2008 6:05 pm Post subject: |
|
|
Quote: |
mrmarky wrote:
3. Look at the clusters greater than 4.
|
Yeah, I meant to put 4 or more. , I've left off this for the minute, working on my generator . I might come back and add some more techniques like this in later.
Thanks
Mark |
|
Back to top |
|
|
|