daj95376 wrote:SteveG48 wrote:Thanks, Danny. JC already pointed out that variant; lovely. However, if that alternative hadn't been available, is there a right way- preferably an elegant one- to do what I was trying to do?
I would work your logic backward and split it into a network.
- Code: Select all
+-----------------------------------------------------------------------+
| 4 156 7 | 156 3 2 | 8 56 9 |
| 15 3 2 | 15689 569 1689 | 46 7 456 |
| 8 569 59 | 456 456 7 | 1 3 2 |
|-----------------------+-----------------------+-----------------------|
| 125 1259 6 | 7 8 49 | 3 129 14 |
| 3 189 89 | 2 469 5 | 46 1689 7 |
| 7 289 4 | 69 1 3 | 5 2689 68 |
|-----------------------+-----------------------+-----------------------|
| 9 2458 1 | 4568 2456 468 | 7 568 3 |
| 25 7 58 | 3 256 168 | 9 4 1568 |
| 6 458 3 | 14589 7 1489 | 2 158 158 |
+-----------------------------------------------------------------------+
# 85 eliminations remain
**************
(6): r8c9 = r7c8 - r1c8 = r2c79 - r2c456 = r3c5 - r8c5
||
= r13c4 - r6c4 = r5c5 - r8c5
**************
The cells bounded by (*) asterisks could be interpreted as a Kraken Box [b2] for <6>.
- Code: Select all
(6): r2c456 - r2c79 = r1c8 - r7c8 = r8c9 - r8c5
(6): r3c5 - r8c5
(6): r13c4 - r6c4 = r5c5 - r8c5
A closer example to your results.
- Code: Select all
**************
(6): r5c5 = r6c4 - r123c4 = r23c5 - r8c5
||
= r2c6 - r2c79 = r1c8 - r7c8 = r8c9 - r8c5
**************
Again, the network could be interpreted as a Kraken Box [b2] for <6>.
Thanks, Danny. Great stuff, as usual. I'll try to absorb it all.
