cannot go on ...

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

cannot go on ...

Postby oberon » Sat Mar 18, 2006 10:00 am

Can you please help me with this AND explain to me how I could have gone on?
Code: Select all
8..|1..|5..
.7.|4..|8..
..9|8.5|...
..3|.24|1..
41.|...|.6.
.9.|...|...
6..|.9.|351
...|.43|.8.
93.|...|724

The puzzle was created with the program "Power sudoku" and as a tip the program suggests a 1 in the 2nd row in the 3rd field - but as far as I can tell it could also be in two other fields of that box:(
greetings from Austria!
oberon
 
Posts: 7
Joined: 18 March 2006

Postby ravel » Sat Mar 18, 2006 12:09 pm

If you are here:
Code: Select all
 *-----------------------------------------------------------------------------*
 | 8       246     246     | 1       367     29      | 5       3479    23679   |
 | 135     7       15      | 4       36      29      | 8       139     2369    |
 | 13      246     9       | 8       367     5       | 246     1347    2367    |
 |-------------------------+-------------------------+-------------------------|
 | 57      568     3       | 5679    2       4       | 1       79      5789    |
 | 4       1       2578    | 3579    58      78      | 29      6       235789  |
 | 257     9       25678   | 3567    158     16      | 24      347     23578   |
 |-------------------------+-------------------------+-------------------------|
 | 6       248     2478    | 27      9       78      | 3       5       1       |
 | 1257    25      1257    | 257     4       3       | 69      8       69      |
 | 9       3       58      | 56      158     16      | 7       2       4       |
 *-----------------------------------------------------------------------------*

You can continue with eliminating the 8 from r5c3 using coloring (what gives you a triple then) - [Edit: Havard pointed out, that only 2 strong links are needed:] cells r4c2,r7c2,r7c6,r5c6.
Alternatively there is a nice loop
r4c2=8=r7c2-8-r9c3-5-r8c2=5=r4c2 => r4c2<>6
That means: if r4c2 is not 8, then r7c2=8 (column 2), it follows r9c3=8 and r8c2=2 and 5 (in column 2) must be in r4c2. So either r4c2=8 or 5.
I cannot see, how r2c3 can be forced to be 1.
Last edited by ravel on Sat Mar 18, 2006 8:51 am, edited 1 time in total.
ravel
 
Posts: 998
Joined: 21 February 2006

Re: cannot go on ...

Postby Havard » Sat Mar 18, 2006 12:25 pm

oberon wrote:Can you please help me with this AND explain to me how I could have gone on?
Code: Select all
8..|1..|5..
.7.|4..|8..
..9|8.5|...
..3|.24|1..
41.|...|.6.
.9.|...|...
6..|.9.|351
...|.43|.8.
93.|...|724

The puzzle was created with the program "Power sudoku" and as a tip the program suggests a 1 in the 2nd row in the 3rd field - but as far as I can tell it could also be in two other fields of that box:(
greetings from Austria!


To be able to crack this puzzle, you will need to have a pretty good grasp on the concept of subsets like pairs and triplets. To find a subset is quite easy, you just have to look for n cells with exactly n different candidates / pencilmarks in the same row, column or box.

valid triplets:
(123) (123) (123) = number of cells: 3 - number of different candidates: 3
(123) (12) (123)
(123) (12) (13)
(12) (13) (23)

not valid triplets:
(123) (1234) (123) = number of cells: 3 - number of different candidates: 4
(12) (145) (123) = number of cells: 3 - number of different candidates: 5

Now the rule goes that if you can find a subset, you can then eliminate all other occurences of the subset-candidates in all other cells in the row, column or box you found the subset in!:)

In this puzzle you will find triplets in box 1 and 2, and a pair in box 9. That should get you started. However, at some point you will need to find a two-strong-link pattern to proceed. Then this link could be helpful:
http://forum.enjoysudoku.com/viewtopic.php?t=3326

havard
Havard
 
Posts: 378
Joined: 25 December 2005

Re: cannot go on ...

Postby Cec » Sat Mar 18, 2006 2:20 pm

oberon wrote:Can you please help me with this AND explain to me how I could have gone on?

Hi oberon,
If you are just starting out you've picked a difficult puzzle. The following may help you get to where ravel started on an advanced technique called "coloring" - more about that later. Your puzzle and initial "candidate" grid ( possible numbers for unsolved cells) would be like this:

Code: Select all
 *-----------*
 |8..|1..|5..|
 |.7.|4..|8..|
 |..9|8.5|...|
 |---+---+---|
 |..3|.24|1..|
 |41.|...|.6.|
 |.9.|...|...|
 |---+---+---|
 |6..|.9.|351|
 |...|.43|.8.|
 |93.|...|724|
 *-----------*

 
 *-----------------------------------------------------------------------------*
 | 8       246     246     | 1       367     2679    | 5       3479    23679   |
 | 1235    7       1256    | 4       36      269     | 8       139     2369    |
 | 123     246     9       | 8       367     5       | 246     1347    2367    |
 |-------------------------+-------------------------+-------------------------|
 | 57      568     3       | 5679    2       4       | 1       79      5789    |
 | 4       1       2578    | 3579    3578    789     | 29      6       235789  |
 | 257     9       25678   | 3567    135678  1678    | 24      347     23578   |
 |-------------------------+-------------------------+-------------------------|
 | 6       248     2478    | 27      9       278     | 3       5       1       |
 | 1257    25      1257    | 2567    4       3       | 69      8       69      |
 | 9       3       158     | 56      1568    168     | 7       2       4       |
 *-----------------------------------------------------------------------------*

Abbreviations are as follows:
r8=row8 c7=column7 L/C means Locked Candidates N/P = naked pair
you refer to "Fields" we call them "boxes" numbered 1 to 9 clockwise top to bottom.

Candidate exclusion as follows:
In r8 N/P [69] in r8c7 and r8c9 leads to exclude 6 from r8c4
In c6 L/C the 2's are locked in box2 exclude 2 from r7c6
In c5 L/C the 3's are locked in box2 exclude 3 from r5c5 and r6c5
In c6 L/C the 9's are locked in box2 exclude 9 from r5c6
In c6 N/P [78} exclude 7's and 8's in other cells in c6
In c6 N/P [16] exclude 6's from other cells in c6
In c5 L/C the 6's are locked in box2 exclude 6's from r6c5 and r9c5
In c5 L/C the 7's are locked in box2 exclude 7's from r5c5 and r6c5

Now you should be at ravel's "coloring" solving technique. I'd suggest you read up and fully understand the basic techniques in the following links before you get into the more advanced techniques which includes "coloring"

http://www.angusj.com/sudoku/hints.php
http://www.simes.clara.co.uk/programs/sudokutechniques.htm

Sorry about all the homework:)

PS Sorry about the "ugly" candidate grid layout - I thought I'd solved this problem. Too late to work on it tonite.

Cec
Cec
 
Posts: 1039
Joined: 16 June 2005

homework

Postby oberon » Sat Mar 18, 2006 3:34 pm

Thanks everybody for helping me - awfully kind, indeed!
Yes I am just starting with the advanced techniques but its time for that as I got a little bored with the easier ones:D
I was quite proud already to find out that my candidate grid was identic to the one you posted!
I am still having difficulties following your suggested loop:

r4c2=8=r7c2-8-r9c3-5...
why is r9c3-5? Sorry?

Anyway I must admit I am really impressed by the developement of sudoku-science - just read the oldest postings in "Advanced solving techniques" ... !:)
oberon
 
Posts: 7
Joined: 18 March 2006

Re: homework

Postby ravel » Sat Mar 18, 2006 4:55 pm

oberon wrote:r4c2=8=r7c2-8-r9c3-5...
why is r9c3-5? Sorry?

r9c3 only has candidates 5 and 8. If r7c2=8, it cannot be 8 (same box).
ravel
 
Posts: 998
Joined: 21 February 2006

Re: homework

Postby oberon » Sat Mar 18, 2006 5:43 pm

ravel wrote:r9c3 only has candidates 5 and 8. If r7c2=8, it cannot be 8 (same box).

Am I getting your syntax wrong?

ravel wrote:r4c2=8=r7c2-8-r9c3-5-r8c2=5=r4c2 => r4c2<>6

If r4c2 is 8 then r7c2 is not 8, meaning it will be 2 or 4, ok?
But then I thought that r9c3 can still be 8 OR 5 because the 8 can also be in r7c3 ??
What am I missing?

[edit: I think now I got it! The 8 in r7c3 can be excluded with the coloring r4c2,r7c2,r7c6,r5c6 the you suggested ...]
oberon
 
Posts: 7
Joined: 18 March 2006

Yesssss!

Postby oberon » Sat Mar 18, 2006 9:36 pm

Thanks to your help I have managed to solve it. Thank you very much!
oberon
 
Posts: 7
Joined: 18 March 2006

Re: cannot go on ...

Postby Cec » Sun Mar 19, 2006 2:26 am

oberon wrote:Can you please help me with this AND explain to me how I could have gone on?
Code: Select all
8..|1..|5..
.7.|4..|8..
..9|8.5|...
..3|.24|1..
41.|...|.6.
.9.|...|...
6..|.9.|351
...|.43|.8.
93.|...|724

Am i missing something that should have "twigged" with me that you had already "gone on" from the puzzle you submitted.? It took me quite a few exclusions as detailed above to assist you to reach ravel's above candidate grid as I assumed you wouldn't have known how to reach that stage - the following however suggests you had already progressed up to the same stage of ravel's grid .
oberon wrote:"..I was quite proud already to find out that my candidate grid was identic to the one you posted!.."

Naturally I'm happy for you to have solved this puzzle but as recommended in the Forum's "Sticky" Thread it does help if you include both your initial puzzle and candidate grids when posting.
Cec
Cec
 
Posts: 1039
Joined: 16 June 2005

Re: cannot go on ...

Postby oberon » Sun Mar 19, 2006 11:51 am

cecbevwr wrote:Am i missing something that should have "twigged" with me that you had already "gone on" from the puzzle you submitted.? It took me quite a few exclusions as detailed above to assist you to reach ravel's above candidate grid as I assumed you wouldn't have known how to reach that stage

To that point I had never solved a sudoku using a candidate grid like you all do - still I did my brainwork (also quite some pencil-and-rubber-work...):) When I wrote down all my considerations in form of such a cg yesterday it looked exactly the same as ravel's.
Evantually reading through all the posted links and mostly because I started using "simple sudoku" it appears that I can meanwhile solve the "simple sudoku"-puzzles in "expert-level" without problems:idea:
oberon
 
Posts: 7
Joined: 18 March 2006

Re: cannot go on ...

Postby Cec » Sun Mar 19, 2006 2:06 pm

Hi oberon,
Oops! My mistake - I'm on my knees for forgiveness as I type this. I overlooked that ravel's grid, whilst eliminating a number of candidates as I did above, still didn't solve any more cells than shown on your initial puzzle. This was an over-sight on my part for which I again apologize.

I too use Simple Sudoku and this Thread may help you understand how to post a puzzle and candidate grid from that program to the forum.
Cec
Cec
 
Posts: 1039
Joined: 16 June 2005


Return to Help with puzzles and solving techniques