The New Sudoku Players' Forum

by **FEARDIZ** » Tue May 02, 2006 6:01 pm

ok I'm confused about colors again..

Code: Select all: *-----------* |.6.|5..|...| |.5.|9..|87.| |.17|...|..3| |---+---+---| |1..|.5.|2..| |.2.|649|.3.| |..4|.1.|..5| |---+---+---| |9..|...|31.| |.41|..7|.8.| |...|..4|.6.| *-----------* *-----------* |.6.|5..|.2.| |.5.|9..|87.| |.17|...|..3| |---+---+---| |1..|.5.|24.| |.2.|649|.3.| |..4|.1.|.95| |---+---+---| |9..|...|31.| |.41|..7|.8.| |...|1.4|.6.| *-----------* *--------------------------------------------------------------------* | 348 6 389 | 5 378 138 | 149 2 149 | | 234 5 23 | 9 236 1236 | 8 7 146 | | 248 1 7 | 248 268 268 | 4569 5 3 | |----------------------+----------------------+----------------------| | 1 3789 3689 | 378 5 38 | 2 4 678 | | 578 2 58 | 6 4 9 | 17 3 178 | | 3678 378 4 | 2378 1 238 | 67 9 5 | |----------------------+----------------------+----------------------| | 9 78 2568 | 28 268 2568 | 3 1 247 | | 2356 4 1 | 23 2369 7 | 59 8 29 | | 23578 378 2358 | 1 2389 4 | 579 6 279 | *--------------------------------------------------------------------*

why can't I exclude 1 from R1C9 ?
I colored like this

Code: Select all: *--------------------------------------------------------------------* | 348 6 389 | 5 378 +138 | -149 2 149 | | 234 5 23 | 9 236 -1236 | 8 7 +146 | | 248 1 7 | 248 268 268 | 4569 5 3 | |----------------------+----------------------+----------------------| | 1 3789 3689 | 378 5 38 | 2 4 678 | | 578 2 58 | 6 4 9 | +17 3 -178 | | 3678 378 4 | 2378 1 238 | 67 9 5 | |----------------------+----------------------+----------------------| | 9 78 2568 | 28 268 2568 | 3 1 247 | | 2356 4 1 | 23 2369 7 | 59 8 29 | | 23578 378 2358 | 1 2389 4 | 579 6 279 | *--------------------------------------------------------------------*

I figured since there is a true and false in R1 it can't be in R1C9
also same thing in C9

thanks

by **Sped** » Tue May 02, 2006 6:26 pm

Code: Select all: *--------------------------------------------------------------------* | 348 6 389 | 5 378 138A | 149X 2 149 | | 234 5 23 | 9 236 1236B | 8 7 146A | | 248 1 7 | 248 268 268 | 4569 5 3 | |----------------------+----------------------+----------------------| | 1 3789 3689 | 378 5 38 | 2 4 678 | | 578 2 58 | 6 4 9 | 17Y 3 178X | | 3678 378 4 | 2378 1 238 | 67 9 5 | |----------------------+----------------------+----------------------| | 9 78 2568 | 28 268 2568 | 3 1 247 | | 2356 4 1 | 23 2369 7 | 59 8 29 | | 23578 378 2358 | 1 2389 4 | 579 6 279 | *--------------------------------------------------------------------*

There is no coloring oportunity on 1s.

r1c6 and r2c6 are a conjugate pair in column 6, as are r2c6 and r2c9 in row 2. I marked them with A and B. r1c7 and r5c7 are a conjugate pair in column 7 as are r5c7 and r5c9 in row 5. I marked them with X and Y. This is multiple coloring, and still doesn't lead to any exclusions.

by **FEARDIZ** » Tue May 02, 2006 7:51 pm

so there is a true and false in the same row or collum or box, how does that not exclude the other numbers?

by **Sped** » Tue May 02, 2006 8:03 pm

FEARDIZ wrote:so there is a true and false in the same row or collum or box, how does that not exclude the other numbers?

In this case there are two true/false chains. A/B and X/Y. Any cell that sees both an A and a B must be false. Any cell that sees both an X and a Y must be false.

But there are no cells like that in the grid.

How did you arrive at r1c6 as a + and r1c7 as a - ??

by **FEARDIZ** » Fri May 05, 2006 6:38 pm

I don't understand multiable colors..
how does the 4 get excluded at R1C6 ?

Code: Select all: *-----------* |9.7|...|..1| |..8|1..|...| |..3|.57|98.| |---+---+---| |1..|.4.|...| |..6|723|4..| |...|.1.|..5| |---+---+---| |.81|96.|3..| |...|..2|1..| |2..|...|8.4| *-----------* *-----------* |9.7|...|..1| |..8|19.|...| |.13|.57|98.| |---+---+---| |1..|.4.|...| |..6|723|41.| |...|.1.|..5| |---+---+---| |.81|96.|3..| |...|..2|1..| |2..|..1|8.4| *-----------* *-----------------------------------------------------------------------------* | 9 2456 7 | 23468 38 468 | 256 246 1 | | 456 2456 8 | 1 9 46 | 2567 23467 2367 | | 46 1 3 | 246 5 7 | 9 8 26 | |-------------------------+-------------------------+-------------------------| | 1 379 29 | 568 4 5689 | 267 23679 236789 | | 58 59 6 | 7 2 3 | 4 1 89 | | 3478 3479 249 | 68 1 689 | 267 23679 5 | |-------------------------+-------------------------+-------------------------| | 47 8 1 | 9 6 45 | 3 257 27 | | 3467 34679 459 | 3458 378 2 | 1 5679 679 | | 2 3679 59 | 35 37 1 | 8 5679 4 | *-----------------------------------------------------------------------------*

by **Sped** » Fri May 05, 2006 8:37 pm

FEARDIZ wrote:I don't understand multiable colors..
how does the 4 get excluded at R1C6 ?

Code: Select all: *-----------------------------------------------------------------------------* | 9 2456 7 | 23468 38 468 | 256 246 1 | | 456 2456 8 | 1 9 46 | 2567 23467 2367 | | 46B 1 3 | 246G 5 7 | 9 8 26 | |-------------------------+-------------------------+-------------------------| | 1 379 29 | 568 4 5689 | 267 23679 236789 | | 58 59 6 | 7 2 3 | 4 1 89 | | 3478 3479 249 | 68 1 689 | 267 23679 5 | |-------------------------+-------------------------+-------------------------| | 47P 8 1 | 9 6 45A | 3 257 27 | | 3467 34679 459 | 3458P 378 2 | 1 5679 679 | | 2 3679 59 | 35 37 1 | 8 5679 4 | *-----------------------------------------------------------------------------*

It looks like you're using Simple Sudoku. Good choice. That is an excellent learning tool.

There are two sets of conjugates on 4s. Blue/Green and Pink/Amber.

Blue and pink share a group in column 1, so if blue is true pink must be false. If pink is false then amber is true. Which leads to.. If blue is true then amber is true.

r1c6 sees a green and an amber. We know that either blue is true or green is true. Any cell that sees green and blue can have 4s excluded. But if blue is true, so is amber. So, any cell that sees green and amber can have also its 4s excluded.

You'll also notice that Pink sees blue and green both. So pink must be false, and amber must be true. Clear 4s from all the pinks and set amber to 4.

It took a while before multiple colors clicked for me, but now I often don't even bother coloring the second set of conjugates.. I just eyeball it and make the exclusions.

Keep using Simple Sudoku. Soon you'll have mastered everything it can throw at you, and you'll be wishing Angus would put some more techniques in it.

by **FEARDIZ** » Fri May 05, 2006 9:49 pm

1st a small wish list for Simple Sudoku ( it's a great tool )
be able to log all moves
be able to copy the coloring into text form

I don't understand how R1C6 gets colored at all, there are other 4's in the box, collum and row

by **Sped** » Fri May 05, 2006 10:12 pm

FEARDIZ wrote:1st a small wish list for Simple Sudoku ( it's a great tool )
be able to log all moves
be able to copy the coloring into text form

I don't understand how R1C6 gets colored at all, there are other 4's in the box, collum and row

R1c6 doesn't get colored. It gets eliminated because it sees both green and amber. Either green or amber must be true, therefore r1c6, which sees both, can't be true.

by **FEARDIZ** » Mon May 08, 2006 5:52 pm

how does it see both or any colors?

by **Sped** » Mon May 08, 2006 6:13 pm

FEARDIZ wrote:how does it see both or any colors?

r1c6 shares box 2 with r3c4, which is green, and it shares column 6 with r7c6, which is amber.

by **FEARDIZ** » Tue May 09, 2006 5:37 pm

ok so does R2C6

and R1C4 sees both colors, so do a few more cells

what make R1C6 special?

by **Sped** » Tue May 09, 2006 10:38 pm

FEARDIZ wrote:ok so does R2C6

and R1C4 sees both colors, so do a few more cells

what make R1C6 special?

Code: Select all: *-----------------------------------------------------------------------------* | 9 2456 7 | 23468 38 468 | 256 246 1 | | 456 2456 8 | 1 9 46 | 2567 23467 2367 | | 46B 1 3 | 246G 5 7 | 9 8 26 | |-------------------------+-------------------------+-------------------------| | 1 379 29 | 568 4 5689 | 267 23679 236789 | | 58 59 6 | 7 2 3 | 4 1 89 | | 3478 3479 249 | 68 1 689 | 267 23679 5 | |-------------------------+-------------------------+-------------------------| | 47P 8 1 | 9 6 45A | 3 257 27 | | 3467 34679 459 | 3458P 378 2 | 1 5679 679 | | 2 3679 59 | 35 37 1 | 8 5679 4 | *-----------------------------------------------------------------------------*

Two cells see green and amber. r1c6 and r2c6. They can both have their 4s excluded.

Also notice that pink sees both blue and green. Pink cannot be true. Exclude 4s from the pinks and set amber to 4. That will take care of r1c6 and r2c6 also.

by **Cec** » Wed May 10, 2006 2:25 am

FEARDIZ wrote:1st a small wish list for Simple Sudoku ( it's a great tool )
be able to log all moves
be able to copy the coloring into text form..."

Hi FEARDIZ,

From my experience when I previously posted a query concerning the SS program it was suggested to me that such queries should be posted to the Non-Pappocom Forum. After doing this the reply I got eradicated my concerns.

I've been following this thread with interest because of Sped's excellent comments on coloring and, if I may make this comment, it appears to me that you have not properly grasped the "basic coloring" concept as previously explained to you in this Thread . I again highlight the following criteria which must be satisfied when applying "basic coloring":

the coloring technique applies to an individual candidate ....... which occupies only two cells in a group (row, column or box) to form a conjugate pair with another candidate in that group."

Here is your grid:

Code: Select all: *--------------------------------------------------------------------* | 348 6 389 | 5 378 138 | 149 2 149 | | 234 5 23 | 9 236 1236 | 8 7 146 | | 248 1 7 | 248 268 268 | 4569 5 3 | |----------------------+----------------------+----------------------| | 1 3789 3689 | 378 5 38 | 2 4 678 | | 578 2 58 | 6 4 9 | 17 3 178 | | 3678 378 4 | 2378 1 238 | 67 9 5 | |----------------------+----------------------+----------------------| | 9 78 2568 | 28 268 2568 | 3 1 247 | | 2356 4 1 | 23 2369 7 | 59 8 29 | | 23578 378 2358 | 1 2389 4 | 579 6 279 | *--------------------------------------------------------------------*

If you look at row 1 where you attempted "coloring" you will see there are three candidate 1's in this row - therefore the "coloring" technique cannot apply for the 1's in this row because the above criteria of only two candidates in a group is not met .

I can well relate to your dilemma in understanding this technique which I struggled with for some time. The "oldies" suggested I thoroughly understood the "Basic" coloring principles before venturing further into the more complex "multiple" coloring technique - I'm glad I took their advice.

Good luck with your coloring

Cec

by **emm** » Wed May 10, 2006 3:35 am

There are two grids in this thread so it's a bit confusing.

These are the candidate 4s in the second grid

Code: Select all: . 4 . | 4 . 4 | . 4 . 4 4 . | . . 4 | . 4 . 4 . . | 4 . . | . . . -------+-------+------ . . . | . . . | . . . . . . | . . . | . . . 4 4 4 | . . . | . . . -------+-------+------ 4 . . | . . 4 | . . . 4 4 4 | 4 . . | . . . . . . | . . . | . . .

For colouring
1. The original cells must be in the same group
2. You can eliminate any candidate that sees both conjugates of the original cells.

These are the links and eliminations

Code: Select all: . 4 . | 4 . 4*| . 4 . 4 4 . | . . 4*| . 4 . 4A. . | 4a. . | . . . -------+-------+------ . . . | . . . | . . . . . . | . . . | . . . 4 4 4 | . . . | . . . -------+-------+------ 4B. . | . . 4b| . . . 4 4 4 | 4B* . | . . . . . . | . . . | . . .

r1c6 and r2c6 see a and b which are the conjugates of the original cells (A and B)

r8c4 also sees a and b – it can be eliminated

r1c4 sees a and B – one original and one conjugate so that doesn’t work...

but here with these links

Code: Select all: . 4 . | 4 . 4 | . 4 . 4 4 . | . . 4 | . 4 . 4a. . | 4A. . | . . . -------+-------+------ . . . | . . . | . . . . . . | . . . | . . . 4 4 4 | . . . | . . . -------+-------+------ 4*. . | . . 4b| . . . 4 4 4 | 4B. . | . . . . . . | . . . | . . .

r7c1 sees a and b and can be eliminated.

by **ravel** » Wed May 10, 2006 11:17 am

Though multi-coloring is a good and very systematic tool, i want to say that in this case the 2 strong links in rows 3 and 7 are easy to spot (old name turbot fish or also advanced coloring with 4 cells).
They lead to the elimination of 4 in r8c4, then r7c6=4 and r12c6<>4. So multi-coloring is not needed here.

The New Sudoku Players' Forum

colors again..

colors again..

New puzzle

Re: New puzzle

Colors again