edit: removed my pencil-mark notation wasn't updated to match your grids just noticed the error.
first grid
found it several ways
a)1678 @ R3C56
b) 1789 @ R3C12,R2C3
C) 1578 @R126C4
RC (ab: 8) (ac:1) Z:7
eliminates:
R2C6 <> 6
030092405420000600005400320002940030700001800060070102309080204200309080000004903
- Code: Select all
+--------------------+---------------------+------------------+
| 168 3 1678 | 1678 9 2 | 4 17 5 |
| 4 2 178 | 1578 135 3578 | 6 179 1789 |
| 1689 1789 5 | 4 (16) 678 | 3 2 1789 |
+--------------------+---------------------+------------------+
| (158) (158) 2 | 9 4 (568) | (57) 3 67 |
| 7 459 34 | 256 235-6 1 | 8 4569 69 |
| 589 6 348 | 58 7 358 | 1 459 2 |
+--------------------+---------------------+------------------+
| 3 157 9 | 1567 8 567 | 2 1567 4 |
| 2 1457 1467 | 3 (156) 9 | (57) 8 167 |
| 1568 1578 1678 | 12567 1256 4 | 9 1567 3 |
+--------------------+---------------------+------------------+
here is an example of the 2-als 2 RC
A) 1567 @ R8C57 <- (2 cells 4 digits)
B)15678 @ R4C1267
C) 16 @ R3C5
RC (A&B) 7
RC (A&C) 1
Z: 6
eliminate
6 @ R5C5
i know hodoku programing swaps the order of ABC around that might be a hiccup slowing it down as order doesn't really matter for xz,xy als and DB
as its uses RC from a-b, b-c, c-a : then interchanges how it allocates abc in its output.
however, order does matter for these versions: first selected als is fixed and all others are subsets of it.
by comparison to the above examplar:
here is what i mean this is from hodoku { these are found from the als-xy finding it notice the set sizes are all n cells with n+1 digits }
Death Blossom: [r8c7], -5- r38c5 {156}, -7- r4c1267 {15678} => r5c5<>6
Almost Locked Set XY-Wing: A=r4c1267 {15678}, B=r38c5 {156}, C=r8c7 {57}, X,Y=5,7, Z=6 => r5c5<>6
since the death blossom is coded using n cells with n+1 digits it wont ever find the 2als- 2rc using the same cells i noted above even though the descriptions for them show that they are supposed to use: almost almost locked sets. {n+2}
