View previous topic :: View next topic 
Author 
Message 
Marty R.
Joined: 12 Feb 2006 Posts: 5501 Location: Rochester, NY, USA

Posted: Mon Apr 26, 2010 9:28 pm Post subject: AIC question 


I feel silly asking this, but those of you who know me know I'm not much of a theorist, and this question has been on my mind. I'm just curious, I have no specific situation that spawned this question.
Are all chains AICs? I can't think of a good antonym for "alternating", but are there chains whose inferences can be strongstrongstrong or weakweakweak? What about things like simple coloring, remote pairs or XYChains? What kind of chains are they? 

Back to top 


Mogulmeister
Joined: 03 May 2007 Posts: 695

Posted: Thu Apr 29, 2010 4:25 pm Post subject: Re: AIC question 


Marty R. wrote: 
Are all chains AICs? I can't think of a good antonym for "alternating", but are there chains whose inferences can be strongstrongstrong or weakweakweak? What about things like simple coloring, remote pairs or XYChains? What kind of chains are they? 
Simple Colouring, XY Chains and remote pairs can all be represented as AICs.
I too am not a great theorist but the following might help  reasonably well written  especially the section on discontinuous Nice Loops.
http://www.paulspages.co.uk/sudokuxp/howtosolve/niceloops.htm 

Back to top 


Marty R.
Joined: 12 Feb 2006 Posts: 5501 Location: Rochester, NY, USA

Posted: Fri Apr 30, 2010 3:43 am Post subject: 


Thanks MM.
Quote:  Simple Colouring, XY Chains and remote pairs can all be represented as AICs. 
If this is true, then I assume that there are no chains that are strongstrongstrong or weakweakweak. 

Back to top 


Mogulmeister
Joined: 03 May 2007 Posts: 695

Posted: Fri Apr 30, 2010 5:11 am Post subject: 


Thats why I directed you to Nice Loops  especially the bit on discontinuities.


Back to top 


wapati
Joined: 10 Jun 2008 Posts: 472 Location: Brampton, Ontario, Canada.

Posted: Fri Apr 30, 2010 5:38 am Post subject: 


Marty R. wrote:  Thanks MM.
Quote:  Simple Colouring, XY Chains and remote pairs can all be represented as AICs. 
If this is true, then I assume that there are no chains that are strongstrongstrong or weakweakweak. 
I am probably lost, I thought all strong links may be considered weak. 

Back to top 


nataraj
Joined: 03 Aug 2007 Posts: 1044 Location: Vienna, Austria

Posted: Fri Apr 30, 2010 6:50 pm Post subject: 


wapati wrote:  Marty R. wrote:  Thanks MM.
Quote:  Simple Colouring, XY Chains and remote pairs can all be represented as AICs. 
If this is true, then I assume that there are no chains that are strongstrongstrong or weakweakweak. 
I am probably lost, I thought all strong links may be considered weak. 
wapati, that is a common misconception. It is NOT true.
strong link means "at least one of the statements must be true"
weak link means: "at most one of the statements is true"
those two are not the same.
In simple cases (most strong links we find at first), like exactly two cells with candidate "6" in a row), there is both a strong link (at least one of the cells must be "6") and a weak link (only one of them can be "6") but in the general case, strong does not imply weak. Take the pincers of a wlink, for example. The wlink means that at least one of the pincers is "x". But it is possible that both are "x". Same with mwing etc.
Chains of weakweak or strongstrong are not particularly useful.
Why?
strong link can also be written as: "if not a then b"
weak link can be written as: "if x then not y"
put those together and you get something new (=better) from the original links:
(if not a then b) and (if b then not c) and (if not c then d) gives: if not a then d. This new relationship can be useful.
two strong links cannot be put together to form a useful new relationship:
(if not a then b) and (if not b then c) do not connect in a useful way
neither can we combine two weak links:
if a then not b and if b then not c does not tell us anything about the relationship between a and c. 

Back to top 


ronk
Joined: 07 May 2006 Posts: 398

Posted: Fri Apr 30, 2010 11:18 pm Post subject: 


wapati wrote:  Marty R. wrote:  Thanks MM.
Quote:  Simple Colouring, XY Chains and remote pairs can all be represented as AICs. 
If this is true, then I assume that there are no chains that are strongstrongstrong or weakweakweak. 
I am probably lost, I thought all strong links may be considered weak. 
I believe you're thinking of a conjugate link, which has both a strong inference and a weak inference. 

Back to top 


wapati
Joined: 10 Jun 2008 Posts: 472 Location: Brampton, Ontario, Canada.

Posted: Sat May 01, 2010 12:11 am Post subject: 


ronk wrote: 
I believe you're thinking of a conjugate link, which has both a strong inference and a weak inference. 
Thank you. 

Back to top 


daj95376
Joined: 23 Aug 2008 Posts: 3855

Posted: Sat May 01, 2010 6:47 pm Post subject: 


[Withdrawn: Probably of little use to Marty.]
Last edited by daj95376 on Sat May 01, 2010 11:20 pm; edited 1 time in total 

Back to top 


ronk
Joined: 07 May 2006 Posts: 398

Posted: Sat May 01, 2010 8:51 pm Post subject: 


daj95376 wrote:  Asellus and ronk are more apt to use an AIC based on (N+1) WIs connected to N SIs. In this case, listing eliminations is redundant because it's a contradiction chain/discontinuous loop where the assumption in the initial WI is disproven by the conclusion in the final WI. 
To be clear, there's no more and no less assumption involved in writing the deduction as a loop ... rather than a chain. Moreover, without a loop there is no deduction. 

Back to top 


Mogulmeister
Joined: 03 May 2007 Posts: 695

Posted: Sat May 01, 2010 10:40 pm Post subject: 


Without being disrespectful to anyone  does this help at all Marty ? 

Back to top 


Marty R.
Joined: 12 Feb 2006 Posts: 5501 Location: Rochester, NY, USA

Posted: Sun May 02, 2010 12:03 am Post subject: 


Mogulmeister wrote:  Without being disrespectful to anyone  does this help at all Marty ? 
I don't want to seem like an ingrate to all who responded, but if my basic question was answered, I missed it. Sorry, but a lot of theory discussion is Greek to me.
Quote:  ...but are there chains whose inferences can be strongstrongstrong or weakweakweak? 


Back to top 


daj95376
Joined: 23 Aug 2008 Posts: 3855

Posted: Sun May 02, 2010 2:02 am Post subject: 


Marty R. wrote:  Quote:  ...but are there chains whose inferences can be strongstrongstrong or weakweakweak? 

Not to my knowledge. 

Back to top 


Marty R.
Joined: 12 Feb 2006 Posts: 5501 Location: Rochester, NY, USA

Posted: Sun May 02, 2010 4:47 am Post subject: 


daj95376 wrote:  Marty R. wrote:  Quote:  ...but are there chains whose inferences can be strongstrongstrong or weakweakweak? 

Not to my knowledge. 
Thanks very much Danny. 

Back to top 


Mogulmeister
Joined: 03 May 2007 Posts: 695

Posted: Sun May 02, 2010 7:38 am Post subject: 


Quote:  but are there chains whose inferences can be strongstrongstrong or weakweakweak? 
Well yes, if you change the candidates. 

Back to top 


nataraj
Joined: 03 Aug 2007 Posts: 1044 Location: Vienna, Austria

Posted: Sun May 02, 2010 1:37 pm Post subject: 


over and out. 

Back to top 


