View previous topic :: View next topic |
Author |
Message |
| AlanLudo
| Joined: 13 Apr 2010 | Posts: 6 | : | | Items |
|
Posted: Tue Apr 13, 2010 4:33 pm Post subject: What are the technics to use |
|
|
Actually, most of the puzzle can be solved with pretty basic technics.
I personally dont go beyond quad.
I have never used swordfish or X-wing
But how to recognise such pattern?
Alan |
|
Back to top |
|
|
| strmckr
| Joined: 24 Apr 2009 | Posts: 52 | : | | Items |
|
Posted: Wed Apr 14, 2010 10:52 pm Post subject: |
|
|
Quote: | Actually, most of the puzzle can be solved with pretty basic techniques. |
not true, 85% of newspaper or distributed puzzles can be solved with a very simplistic solving techniques.
there are far more puzzles that require beyond these,
for example:
platnum blonde
.......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6...
Code: | .----------------------.------------------------.-----------------------.
| 35678 34589 345679 | 4679 5689 4578 | 679 1 2 |
| 15678 14589 45679 | 124679 125689 124578 | 679 56789 3 |
| 15678 1589 2 | 3 15689 1578 | 4 56789 6789 |
:----------------------+------------------------+-----------------------:
| 237 2349 1 | 8 236 234 | 23679 234679 5 |
| 235 6 3459 | 124 7 12345 | 8 2349 149 |
| 23578 23458 3457 | 1246 12356 9 | 12367 23467 1467 |
:----------------------+------------------------+-----------------------:
| 1236 123 8 | 5 1239 1237 | 123679 234679 14679 |
| 9 123 36 | 127 4 12378 | 5 23678 1678 |
| 4 7 35 | 129 12389 6 | 1239 2389 189 |
'----------------------'------------------------'-----------------------' |
however they are never printed due to the level of frustration they will invoke to an average solver. coupled with the length of time to solve also becomes impractical to solver via a commute ride.
Quote: |
I have never used swordfish or X-wing
But how to recognise such pattern? |
use pm's for the single digit only then look for the patterns |
|
Back to top |
|
|
| AlanLudo
| Joined: 13 Apr 2010 | Posts: 6 | : | | Items |
|
Posted: Thu Apr 15, 2010 12:32 am Post subject: |
|
|
Yep, that was actually something I noticed as the solver I developed is using quads as theist difficult solving technic and so far none of the puzzle I found on news paper or so have been insolvable by my engine.
I didn't precisely get what you wanted to say about the swordfish pattern recognition ?
Cheers,
Alan |
|
Back to top |
|
|
| strmckr
| Joined: 24 Apr 2009 | Posts: 52 | : | | Items |
|
Posted: Thu Apr 15, 2010 1:35 am Post subject: |
|
|
here it is for x-wings, the same idea is for NxN fish constraints.
two rows / columns where there is only 4 digits.
the digits of N form a rectangle shape.
where each digit see one digit on the opposite row.
the eliminations occur in opposite direction to where the digits line up.
( if its two rows. eliminate on columns,
if its two columns eliminate on rows)
to understand how the elimination works
try placing a N on the board, the opposite corner is always true.
thus no matter which way the rectangle solves the x's are never true
Code: | / / N | / / / | / / N
. . x | . . . | . . x
. . x | . . . | . . x
--------+-------+-------
/ / N | / / / | / / N
. . x | . . . | . . x
. . x | . . . | . . x
--------+-------+-------
. . x | . . . | . . x
. . x | . . . | . . x
. . x | . . . | . . x |
Quote: | KEY: '/' <=> cells void of candidate 'N'
'N' <=> cells with candidate 'N'; all are required
"-X" <=> potential eliminations of candidate 'N' |
Example on Rows
digit 1
on Rows 4 and 7 there is only 2 locations per row where the digits 1 are
they form a rectangle shape.
eliminate all the ones on the columns.
Code: | .-----------------.-------------------------.--------------------.
| 1 1346 2 | 146 46-1 7 | 9 5 8 |
| 7 456 4589 | 45689 45689 4689 | 1 2 3 |
| 189 1456 4589 | 145689 245689-1 3 | 467 467 47 |
:-----------------+-------------------------+--------------------:
| 5 9 48 | 7 1248@ 1248@ | 3 48 6 |
| 128 147 478 | 14689 234689-1 24689-1| 478 4789 4579 |
| 3 47 6 | 489 489 5 | 478 1 2 |
:-----------------+-------------------------+--------------------:
| 9 57 579 | 2 1456789@ 14689@ | 4678 346789 479 |
| 6 8 3 | 459 4579 49 | 247 479 1 |
| 4 2 1 | 3 6789 689 | 5 6789 79 |
'-----------------'-------------------------'--------------------' |
1 r47 c56 => r135c5,r5c6<>1 |
|
Back to top |
|
|
| champagne
| Joined: 14 Apr 2010 | Posts: 6 | : | | Items |
|
Posted: Fri Apr 16, 2010 11:45 am Post subject: |
|
|
strmckr wrote: |
for example:
platinum blonde
.......12........3..23..4....18....5.6..7.8.......9.....85.....9...4.5..47...6...
however they are never printed due to the level of frustration they will invoke to an average solver. coupled with the length of time to solve also becomes impractical to solver via a commute ride.
|
Hi strmckr,
I was just prepared to publish a "relatively short" solution to Platinium Blonde combining "ttt"s findings and my solver path (before I introduce changes in the solver to be in a position to produce such a path) when the player's forum vanished.
I understand from others that I have first to post 5 messages before I can use links that are necesssary to do it.
I start now and I will continue if you are interested.
champagne |
|
Back to top |
|
|
| strmckr
| Joined: 24 Apr 2009 | Posts: 52 | : | | Items |
|
Posted: Sat Apr 17, 2010 1:18 am Post subject: |
|
|
Quote: | I was just prepared to publish a "relatively short" solution to Platinium Blonde combining "ttt"s findings and my solver path (before I introduce changes in the solver to be in a position to produce such a path) when the player's forum vanished. |
a short solution outside my own would be interesting to see, however the puzzle was used merely as an example of something more complex ie outside the newspaper printed puzzle. |
|
Back to top |
|
|
| champagne
| Joined: 14 Apr 2010 | Posts: 6 | : | | Items |
|
Posted: Sat Apr 17, 2010 4:46 am Post subject: |
|
|
strmckr wrote: |
a short solution outside my own would be interesting to see, however the puzzle was used merely as an example of something more complex ie outside the newspaper printed puzzle. |
I don't know what is your solution. I only know for that puzzle what has been done by "ttt" and "Allan Barker".
The path I am working on is using intensively Aur's (the start of ttt) and virus chains.
where was published your solution??
champagne |
|
Back to top |
|
|
| strmckr
| Joined: 24 Apr 2009 | Posts: 52 | : | | Items |
|
Posted: Mon Apr 19, 2010 12:20 am Post subject: |
|
|
no i didn't post my findings:
the most i've done with the monster puzzles is identify interesting virus type patterns
here by Allan confirmed there existence with his program (i've only done it to a few of the monsters)
i'll see if i can remember exactly what i used in solving and post it
{i lost a lot of my work on a reformat after my computer was hacked}
cheers
strmckr. |
|
Back to top |
|
|
|