help: mult colors

Post the puzzle or solving technique that's causing you trouble and someone will help

help: mult colors

Postby mikebot » Sun Jul 30, 2006 11:27 pm

can someone type this game into simple sudoku:
Code: Select all
382 6*1 9*5
94* 8*3 *2*
6** 2** 8*3

19* 5** 234
7*4 *2* 158
25* **4 *9*

5** **8 4*2
42* 7*5 389
8*9 4*2 56*


and press 'f7', and explain to me how you can remove that '1' based on multi. colors? thanks.
mikebot
 
Posts: 4
Joined: 13 May 2006

Postby Sped » Mon Jul 31, 2006 2:00 am

Code: Select all
 
 *-----------------------------------------------------------*
 | 3     8     2     | 6     47    1     | 9     47    5     |
 | 9     4     157A  | 8     57    3     | 67    2     167a  |
 | 6     17    157   | 2     4579  79    | 8     147A  3     |
 |-------------------+-------------------+-------------------|
 | 1     9     68    | 5     78    67    | 2     3     4     |
 | 7     36    4     | 39    2     69    | 1     5     8     |
 | 2     5     38    | 13    138   4     | 67    9     67    |
 |-------------------+-------------------+-------------------|
 | 5     1367 (1)367 | 139   1369  8     | 4     17a   2     |
 | 4     2     16    | 7     16    5     | 3     8     9     |
 | 8     137   9     | 4     13    2     | 5     6     17    |
 *-----------------------------------------------------------*


It's simple colors, not multi colors.

The conjugate 1s in row 2 are marked "A" and "a". Conjugate in this case means that there are just two 1s in the row such that one must be true for 1 and the other false.

In box 3 the conjugate 1s are marked "a" (r2c9) and "A" (r3c8). Either all the "A" cells are 1 or all the "a" cells are 1.

The conjugate 1s in column 8 are r3c8 "A" and r7c8 "a".

It so happens that r7c3 sees both the "A" in r2c3 and the "a" in r7c8.

Since either all the "A" cells are 1 or all the "a" cells are 1, we know that a cell that sees both "a" and "A" cannot be 1.

Therefore the 1 can be excluded from r7c3.
Sped
 
Posts: 126
Joined: 26 March 2006

Postby tso » Mon Jul 31, 2006 5:43 am

It's the 1 from r7c4 that Mikebot is asking about.


Code: Select all
 *-----------------------------------------------------------*
 | 3     8     2     | 6     47    1     | 9     47    5     |
 | 9     4    +157   | 8     57    3     | 67    2    -167   |
 | 6     17    157   | 2     4579  79    | 8    +147   3     |
 |-------------------+-------------------+-------------------|
 | 1     9     68    | 5     78    67    | 2     3     4     |
 | 7     36    4     | 39    2     69    | 1     5     8     |
 | 2     5     38    | 13    18    4     | 67    9     67    |
 |-------------------+-------------------+-------------------|
 | 5     1367  367   |x19    1369  8     | 4    -17    2     |
 | 4     2    ^16    | 7    v16    5     | 3     8     9     |
 | 8     137   9     | 4     13    2     | 5     6     17    |
 *-----------------------------------------------------------*


If r2c3<>1, follow the chain clockwise:

r2c3<>1 -> r2c9=1 -> r3c8<>1 -> r7c8=1 -> r7c4<>1

If r2c3=1, follow it counter-clockwise.

r2c3=1 -> r8c3<>1 -> r8c5=1 -> r7c4<>1
tso
 
Posts: 798
Joined: 22 June 2005

Postby emm » Mon Jul 31, 2006 5:53 am

Here’s how the chain is explained by multiple colouring.

You have two different chains - labelled Aa and Bb - where the first cell of each chain (A & B) share a group

Code: Select all
 . . . | . . . | . . .
 . . 1A| . . . | . . 1a
 . 1 1 | . . . | . 1A.
-------+-------+------
 . . . | . . . | . . .
 . . . | . . . | . . .
 . . . | 1 1 . | . . .
-------+-------+------
 . 1 . | 1*1 . | . 1a.
 . . 1B| . 1b. | . . .
 . 1 . | . 1 . | . . 1A


A and B share a group so they can’t both be true so either a or b are true and any cell that shares a group with both a and b is false ie r7c4

This can also be visualised with strong links - see this topic

You can remove 1s from both r7c4 and r7c5 as they see both ends of the links at r7c8 and r8c5

Code: Select all
 . . . | . . . | . . .
 . . 1-|-------|-----1
 . 1 1 | . . . | . 1 .
-------+-------+---|---
 . . . | . . . | . | .
 . . . | . . . | . | .
 . . . | 1 1 . | . | . 
-------+-------+---|--
 . 1 . | 1*1*. | . 1 .
 . . 1-|---1 . | . . .
 . 1 . | . 1 . | . . 1
emm
 
Posts: 987
Joined: 02 July 2005

Does SS do UR's??

Postby keith » Mon Jul 31, 2006 6:58 am

Why not just use the UR's to eliminate <7> from R3C5,8, and also <67> from R2C9?

Then, an XY-wing solves the puzzle.

Is it that Simple Sudoku does not do Unique Rectangles?

Keith

Code: Select all
+----------------+----------------+----------------+
| 3    8    2    | 6    47   1    | 9    47   5    |
| 9    4    157  | 8    57   3    | 67   2    167  |
| 6    17   157  | 2    4579 79   | 8    147  3    |
+----------------+----------------+----------------+
| 1    9    68   | 5    78   67   | 2    3    4    |
| 7    36   4    | 39   2    69   | 1    5    8    |
| 2    5    38   | 13   138  4    | 67   9    67   |
+----------------+----------------+----------------+
| 5    1367 1367 | 139  1369 8    | 4    17   2    |
| 4    2    16   | 7    16   5    | 3    8    9    |
| 8    137  9    | 4    13   2    | 5    6    17   |
+----------------+----------------+----------------+
keith
2017 Supporter
 
Posts: 221
Joined: 03 April 2006

Re: Does SS do UR's??

Postby udosuk » Mon Jul 31, 2006 8:05 am

keith wrote:Is it that Simple Sudoku does not do Unique Rectangles?

Simple Sudoku doesn't do Unique Rectangles... If it does than all of us who use it to solve killer sudoku, sudoku X or other variants will be in big trouble...:D
udosuk
 
Posts: 2698
Joined: 17 July 2005

Postby daj95376 » Mon Jul 31, 2006 9:04 am

I probably have an incorrect understanding on how Multi-Coloring works, but I'm going to present it anyway.

Below should be the grid where you encounter Multi-Coloring. The hint is to set [r7c4]<>1. With my interpretation, [r7c5]<>1 should also be indicated, but it isn't ... and it isn't critical.

Use Blue/Green to indicate the conjugate pairs chain in [r2], [b3], and [b9] for <1>. This represents an either/or mapping on where <1> can be a solution in these cells. Use Pink/Amber to indicate the conjugate pair in [r8] for <1>. Again, an either/or relationship.

Now, the key to understanding Multi-Coloring in this grid is to realize that [c3] has two ends of different chains in common. Since Blue and Pink can not both be <1>, this translates into Blue and Amber having an equivalent relationship to Pink. I now treat Blue and Amber as being equivalent. This being the case, then any <1> in the overlap of Amber and Green can be removed. Thus, [r7c45]<>1. From here, the puzzle is solved with Singles.

Code: Select all
 *-----------------------------------------------------------*
 | 3     8     2     | 6     47    1     | 9     47    5     |
 | 9     4     157B  | 8     57    3     | 67    2     167G  |
 | 6     17   ?157   | 2     4579  79    | 8     147B  3     |
 |-------------------+-------------------+-------------------|
 | 1     9     68    | 5     78    67    | 2     3     4     |
 | 7     36    4     | 39    2     69    | 1     5     8     |
 | 2     5     38    | 13    18    4     | 67    9     67    |
 |-------------------+-------------------+-------------------|
 | 5     1367  367   |*19   *1369  8     | 4     17G   2     |
 | 4     2     16P   | 7     16A   5     | 3     8     9     |
 | 8     137   9     | 4     13    2     | 5     6     17B   |
 *-----------------------------------------------------------*

Note: It's tempting to think of Pink and Green as being equivalent. If they are equivalent, then [r3c3]<>1 can also be deduced.
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby ravel » Mon Jul 31, 2006 9:35 am

daj95376 wrote:Note: It's tempting to think of Pink and Green as being equivalent. If they are equivalent, then [r3c3]<>1 can also be deduced.
No, there is still a valid template with r3c3=1:
Code: Select all
 . . . | . . 1 | . . .
 . . . | . . . | . . 1
 . . 1 | . . . | . . .
-----------------------
 1 . . | . . . | . . .
 . . . | . . . | 1 . .
 . . . | 1 . . | . . .
-----------------------
 . . . | . . . | . 1 .
 . . . | . 1 . | . . .
 . 1 . | . . . | . . .
Since blue and pink share a unit, they cannot both be true. That means, that either green or amber (or both) must be true.
But green and amber dont share a unit.
ravel
 
Posts: 998
Joined: 21 February 2006

Postby udosuk » Mon Jul 31, 2006 12:07 pm

ravel wrote:Since blue and pink share a unit, they cannot both be true...

Blue and pink cannot both be true, but they can be both false, which enables r3c3 to be 1.

Green and amber cannot be both false, so either one of them must be true, and also they can be both true (since there were no interaction between their cells).

Blue and amber are not equivalent... You can only conclude that if blue is true, amber must be true... If blue is false, amber can be true or false... So the true relationship is "blue => amber" instead... And the relationship between pink and green is if pink is true then blue must be false and green must be true, i.e. "pink => green"...

And if you look at it the other way it was the possibility of r3c3 being 1 that forced you to label r8c35 with another set of colours instead of using simple colouring in the first place...:)
udosuk
 
Posts: 2698
Joined: 17 July 2005

Postby daj95376 » Mon Jul 31, 2006 5:05 pm

Thanks for the clarification ravel & udosuk !!!

So, any cell in the overlap of the non-linked colors -- Amber and Green in this case -- can have the designated candidate eliminated. At least, this still means that <1> can be eliminated from [r7c45].
daj95376
2014 Supporter
 
Posts: 2624
Joined: 15 May 2006

Postby mikebot » Mon Jul 31, 2006 10:18 pm

thanks a lot for the help everyone, i get it now. it is way harder to find multi colors than it is to get them once you see them.
mikebot
 
Posts: 4
Joined: 13 May 2006


Return to Help with puzzles and solving techniques