coloin wrote:Applying the SK loop will reduce the 1-template counts - and hence the n-templates generally.
So i think ronk is asking how many 4-templates for digits <1267> are there before and after applying SK loop eliminations.
My question was poorly phrased, likely because of fuzzy thinking. What I meant was, adopting
daj95376's notation ... are there exclusions based on the <1267>-templates over and above exclusions based on the
sk-loop, aka
hidden-pair-loop?
daj95376 wrote:Here is the <1267>-rookery. Of 524,472 combinations, only 23 survived. I couldn't apply an sk-loop for comparison.
- Code: Select all
.............VV...................V..V.....V..V...................VV.............
17.6....22...17..6..62..71...1.6.27..2..71.6..67..21..71....62.6..12...7..27.6..1
17.6....22...17..6..62..71...7.2.16..1..76.2.62...1..77..1.26....176.2...62....71
17.6....22...17..6..62..71.....2.167.1..76.2.627..1...7..1.26....176.2...62....71
17...6..22...17..6..62..71...7.2.16..1.67..2.62...1..77..1.26....176.2...62....71
17...6..22...17..6..62..71.....2.167.1.67..2.627..1...7..1.26....176.2...62....71
1..7.6..2..7.12..6.26...71.2...6.17..1.27..6.67...12..7..1..62...162...7.62..7..1
17.6....22...17..6..6..271.6...2.17..2.17..6..17..62..7....162...126...7.627....1
17.6....22...17..6..6..271.6...2.17..2..71.6..17..62..7..1..62...126...7.627....1
1..6.7..2..7.21..6.26...71.6...1.27..1..72.6.27...61..7..1..62...126...7.627....1
1..6.7..2..7.21..6.26...71.6...1..27.1.27..6.27...61..7..1.26....176.2...62....71
17.....622...671....61.27....1.2..76.2.67..1..67..12..71....62.6..21...7..27.6..1
1..7...62..7.261...261..7..2.1.6..7..6..72.1..7...12.671....62.6..21...7..26.7..1
1..7...62..7.261...261..7....1.6.27..6..72.1.27...1..671....62.6..21...7..26.7..1
1....7.62..7.261...261..7....1.6..27.6.27..1.27...1..671...26..6..71.2....26...71
1....7.62..7.261...261..7....1.6..27.6..72.1.27...1..671.2..6..6..71.2....26...71
17.....622...671....61.27....7.2..16.6..71.2..21..6..771.2..6..6..71.2....26...71
1....7.62..7.261...261..7..2...6..17.6..71.2..71..2..671.2..6..6..71.2....26...71
17.....622...671....62.17..6...1.27..2..76.1..17..2..67..1..62...162...7.627....1
17.....622...671....62.17......1.276.2..76.1.617..2...7..1..62...162...7.627....1
1..7...62..7.621...26..17..2...1..76.6.27..1..71..62..71....62.6..12...7..26.7..1
1..7...62..7.621...26..17......1.276.6.27..1.271..6...71....62.6..12...7..26.7..1
17.....622...671....62.17....7.2..16.6.17..2..21..6..771...26..6..71.2....26...71
1....7.62..7.261...26..17..2...6..17.6.17..2..71..2..671.2..6..6..71.2....26...71
I added a "V" header line which marks the 8 cells limited to "sk-loop-digits", aka <1267>-template digits. In grid format, this looks like ...
- Code: Select all
. . . | . . . | . . .
. . . | . V V | . . .
. . . | . . . | . . .
-------+-------+-------
. . . | . . . | . V .
. V . | . . . | . V .
. V . | . . . | . . .
-------+-------+-------
. . . | . . . | . . .
. . . | V V . | . . .
. . . | . . . | . . .
Therefore, we have r28c5,r5c28=126 and r2c6,r4c8,r6c2,r8c2=1267 with the consequential exclusion of other digits in those cells.
There are other exclusions however, some part of the sk-loop technique, and some not. As an example of the former, either r1c2=7 or r2c3=7. Templating should eliminate any 7s candidate that sees both of these. As an example of the latter, r4c5=126 for exclusion r4c5<>4, which AFAIK is beyond the scope of the sk-loop.