SudoCue Users

Posted: **Wed Apr 18, 2007 8:24 pm**

[Edited to correct error in start position noted by Mike below]

Start position for the 17 April 2007 Daily Nightmare

. . 7|. . 5|. 2 .
. 2 .|. 6 .|. 5 8
4 . .|2 . .|. . .
-----+-----+-----
. . 2|. 5 .|. . 9
. 1 .|6 . 7|. 4 .
5 . .|. 4 .|3 . .
-----+-----+-----
. . .|. . 9|. . 6
1 6 .|. 3 .|. 7 .
. 8 .|4 . .|1 . .

This is the position after initial basic eliminations:

Code: Select all

.------------------.-------------------.------------------.
| 68    39    7    | 1389   189   5    | 469   2     134  |
| 39    2     1    |*379    6     4    | 79    5     8    |
| 4     5     68   | 2     #1789  138  | 679   1369 *137  |
&#58;------------------+-------------------+------------------&#58;
| 678   4     2    | 138    5     138  | 678   168   9    |
| 389   1     389  | 6      289   7    | 258   4     25   |
| 5    *79    689  | 189    4     128  | 3     168  #127  |
&#58;------------------+-------------------+------------------&#58;
| 237  #37    345  | 158-7  1278  9    | 2458  38    6    |
| 1     6     459  | 58     3     28   | 24589 7     245  |
| 2379  8     359  | 4      27    6    | 1     39    235  |
'------------------'-------------------'------------------'

The following AIC for digit 7 is enough to complete the puzzle with only singles thereafter. (Cells in the chain are alternately marked with "*" and "#" in the diagram.)

(7): r2c4 = r3c5 - r3c9 = r6c9 - r6c2 = r7c2 => r7c4 <> 7

Posted: **Wed Apr 18, 2007 9:23 pm**

Hi

I generally prefer looking in patterns when solving sudoku's.
This can be recreated with this Finned Swordfish on 7 in C259 with R9C5 being the fin.

Code: Select all

.------------------.-------------------.------------------.
| 68    39    7    | 1389   189   5    | 469   2     134  |
| 39    2     1    | 379    6     4    | 79    5     8    |
| 4     5     68   | 2     *1789  138  | 679   1369 *137  |
&#58;------------------+-------------------+------------------&#58;
| 678   4     2    | 138    5     138  | 678   168   9    |
| 389   1     389  | 6      289   7    | 258   4     25   |
| 5    *79    689  | 189    4     128  | 3     168  *127  |
&#58;------------------+-------------------+------------------&#58;
| 237  *37    345  | 158-7 *1278  9    | 2458  38    6    |
| 1     6     459  | 58     3     28   | 24589 7     245  |
| 2379  8     359  | 4     #27    6    | 1     39    235  |
'------------------'-------------------'------------------'

greetings

Para

Posted: **Thu Apr 19, 2007 7:19 pm**

As I understand it, Para's finned Swordfish would actually be equivalent to the AIC:

(7): r79c5 = r3c5 - r3c9 = r6c9 - r6c2 = r7c2 => r7c4 <> 7

where the fin belongs to the grouped node r79c5.

By the way, Ron, your "Start position" grid is incorrect in the top band. Maybe you can quickly fix that? Thanks.

Posted: **Tue Apr 24, 2007 1:25 am**

Original post edited to correct the error Mike spotted -- thanks.

Yes, there's more than one way to view this. Grouped AIC's like the one Mike gives can always be written for empty rectangle and finned X-wing eliminations, and for finned 2-2-2 swordfish patterns like this one.

SudoCue Users

17 April 2007

17 April 2007

Re: 17 April 2007