JC, Danny and Norm: Thank you for the comments.
JC: you had me going with the "Pointing" description, until I realized that Ronk's "Fig 2B" corresponds to the case that I was using, and the eliminations are equivalent to the ones from "pointing" locked candidates in "b5 intersect c5".
The short answer to my question, seems to be "It's a Franken Fish, even if boxes only appear as the cover sectors".
I was interested in this case too, which seems to be missing above:
- Code: Select all
Unlisted (?) cc\rb variation:
. . . | X * X | . . .
. . . | X * X | . . .
. . . | X * X | . . .
---------+----------+----------
. . . | / . / | . . .
. . . | / . / | . . .
. . . | / . / | . . .
---------+----------+----------
* * * | X * X | * * *
. . . | / . / | . . .
. . . | / . / | . .
--
I should say why I proposed my particular "solution".
It was to address some of the issues that Danny has been looking at.
--
This is the diagram Danny's original 5-SIS elimination for 2r4c2.
- Code: Select all
+------------------+-----------+---------------------+
| 2348 1 234 | 368 7 9 | 238(6) 5 2468 |
| 9 38 5 | 368 4 2 | 7 1 (68) |
| 2348 7 6 | 38 1 5 | 238 2348 9 |
+------------------+-----------+---------------------+
| 234 3-2 7 | 1 5 8 | 9 2346 246 |
| 1 6 348 | 9 2 7 | 38 348 5 |
| 5 9 8(2) | 4 3 6 | 1 8(2) 7 |
+------------------+-----------+---------------------+
| 238 5 23 | 7 6 1 | 4 9 (28) |
| 6 4 9 | 2 8 3 | 5 7 1 |
| 7 8(2) 1 | 5 9 4 | 8(6-2) 68(2) 3 |
+------------------+-----------+---------------------+
(Translated and) entered into XSudo, it gives:
5 Truths = {2R69 6C7 27N9}
7 Links = {2c28 8c9 9n7 2b49 6b3}
2 Eliminations --> r4c2<>2, r9c7<>2
It shows a cannibal elimination for 2r9c7, (presumably) based on this "embedded" (discounuous) loop:
"2r9c7 - 6r9c7 = r1c7 - (6=8)r2c9 - (8=2)r7c9 - 2r9c7"
--
This is for an alternate (6-SIS approach), based on Danny's suggestion to use "2r7c9 - r7c13 = 2r9c2" with his "oddagon" based network.
It also corresponds to Norm's "shortened" version.
- Code: Select all
+-------------------+-----------+---------------------+
| 2348 1 234 | 368 7 9 | 238(6) 5 2468 |
| 9 38 5 | 368 4 2 | 7 1 (68) |
| 2348 7 6 | 38 1 5 | 238 2348 9 |
+-------------------+-----------+---------------------+
| 234 3-2 7 | 1 5 8 | 9 2346 246 |
| 1 6 348 | 9 2 7 | 38 348 5 |
| 5 9 8(2) | 4 3 6 | 1 8(2) 7 |
+-------------------+-----------+---------------------+
| 38(2) 5 3(2) | 7 6 1 | 4 9 (28) |
| 6 4 9 | 2 8 3 | 5 7 1 |
| 7 8(2) 1 | 5 9 4 | 8(6-2) 68(2) 3 |
+-------------------+-----------+---------------------+
(Translated and) entered into XSudo, it gives:
6 Truths = {2R69 6C7 27N9 2B7}
7 Links = {2r7 2c28 8c9 9n7 2b4 6b3}
2 Eliminations --> r4c2<>2, r9c7<>2
It shows a cannibal elimination for 2r9c7 again, (presumably) based on the presence of this "embedded" (discounuous) loop:
"2r9c7 - 6r9c7 = r1c7 - (6=8)r2c9 - (8=2)r7c9 - 2r7c13 = 2r9c2 - 2r9c7"
In this case, the final weak link would come on account of 2R9 being one of the "truth" sets, and XSudo treating its presence as (also) specifying a weak link: "2r9c2 - 2r9c7".
--
This is for my alternate (7-SIS) elimination for 2r4c2).
- Code: Select all
+---------------------+-----------+---------------------+
| 348(2) 1 34(2) | 368 7 9 | 238(6) 5 2468 |
| 9 38 5 | 368 4 2 | 7 1 (68) |
| 348(2) 7 6 | 38 1 5 | 238 2348 9 |
+---------------------+-----------+---------------------+
| 34(2) 3-2 7 | 1 5 8 | 9 2346 246 |
| 1 6 348 | 9 2 7 | 38 348 5 |
| 5 9 8(2) | 4 3 6 | 1 8(2) 7 |
+---------------------+-----------+---------------------+
| 38(2) 5 3(2) | 7 6 1 | 4 9 (28) |
| 6 4 9 | 2 8 3 | 5 7 1 |
| 7 8(2) 1 | 5 9 4 | 8(26) 68(2) 3 |
+---------------------+-----------+---------------------+
(Translated and) entered into XSudo, it gives:
7 Truths = {2R69 2C13 6C7 27N9}
8 Links = {2r7 2c28 8c9 9n7 2b14 6b3}
1 Elimination --> r4c2<>2
Note: No cannibal eliminations were detected !
[ I have to run ... I hope I didn't include any typos above ... Cheers ! ]
