The New Sudoku Players' Forum

[daj95376]

Below, I took my solver's Template output and noticed a simple Double Implication Forcing Chains. (This is nothing new for me.) From there, I noticed a strong/weak links relationship. (This is new for me.) From there, I created a simple chain/Nice Loop. (This is very new for me.) I'm beginning to understand why others take my DIFC output and turn it into chains/loops. They are more compact -- even if they do conceal my original thought process by burying the DIFC in the middle of the chain.

Code: Select all

 *-----------*
 |7..|.1.|..5|
 |..8|5.4|6..|
 |...|...|...|
 |---+---+---|
 |.3.|.8.|.6.|
 |4..|...|..7|
 |5.6|...|2.3|
 |---+---+---|
 |...|.9.|...|
 |.7.|3.8|.2.|
 |.1.|247|.9.|
 *-----------* # puzzle "This one has lots." from wapati

# before SSTS does Colors on <5>
 *--------------------------------------------------------------------*
 | 7      46     2349   | 8      1      2369   | 349    34     5      |
 | 139    29     8      | 5      23     4      | 6      7      129    |
 | 369    456    13459  | 7      236    369    | 3489   1348   2489   |
 |----------------------+----------------------+----------------------|
 | 129    3      7      | 49     8      25     | 1459   6      149    |
 | 4      289    19     | 6     -235    35     | 589   ~158    7      |
 | 5      89     6      | 49     7      1      | 2      48     3      |
 |----------------------+----------------------+----------------------|
 | 2368   456    2345   | 1      9     -56     | 7     ~3458   468    |
 | 69     7     -459    | 3      56     8      |-145    2      146    |
 | 368    1      35     | 2      4      7      | 358    9      68     |
 *--------------------------------------------------------------------*


The above eliminations on <5> can be performed in several ways. Templates and a Restricted DIFC (on [r57c8] using only <5>) will do them all at once. Simple Sudoku performs these eliminations in several steps. Finally, the following strong/weak links can be represented as a simple chain.

Code: Select all

[r8c5]=5=[r5c5]-5-[r5c8]
[r8c5]=5=[r7c6]-5-[r7c8]

[r8c5]=5=[r7c6]-5-[r7c8]=5=[r5c8]-5-[r5c5]=5=[r8c5]


Thus, [r8c5]=5 and all of the eliminations are performed.

Question #1: Why would anyone object to having a simple chain like this in a solution?

Question #2: Is my chain really a Nice Loop? (reading the thread on NLs makes me dizzy.)

Question #3: When a DIFC is restricted to tracking only one value, is there any objection to my labeling it a Restricted DIFC?

[ravel]

This is a closed chain with 3 strong links, what means, "r5c8 or r5c8 must be 5" .
You can find it, when you (try to) combine several strong links for eliminations.

[Added as my contribution to the confusions caused by many names for the same:]For the nostalgics: This is a turbot fish type 2 and contains 2 turbot fishes of type 4.

[wapati]

Question #1: Why would anyone object to having a simple chain like this in a solution?

If you were aiming at people like me...I don't use chains or understand the way they are written. I could learn it, I could learn colors.

I don't wish to. I like patterns.

I am amazed at what some chains can do to reduce a puzzle.
I respect them, admire them, I don't want to do them, as yet.

[ronk]

Question #3: When a DIFC is restricted to tracking only one value, is there any objection to my labeling it a Restricted DIFC?

"Restricted" is a general term and subject to misinterpretation. I suggest being specific with something like "1-digit DIFC."

Your nice loop expression is correct. BTW what's the essential difference between DIC and DIFC?

[Carcul]

Question #1: Why would anyone object to having a simple chain like this in a solution?

I don't think there is any reason to object that.

Question #2: Is my chain really a Nice Loop? (reading the thread on NLs makes me dizzy.)

Yes, more precisely a discontinuous Nice Loop.

Question #3: When a DIFC is restricted to tracking only one value, is there any objection to my labeling it a Restricted DIFC?

I think you can call it whatever you want. But the question is: is that really necessary?

Carcul

[daj95376]

Thanks ravel, ronk, and Carcul for the feedback!

ravel: I see your point. BTW, do you have a type?

ronk: I believe the original name was Double Implication Chains (DIC). However, I read a post recently where the phrase Double Implication Forcing Chains (DIFC) was used. Although I don't remember who wrote it, I remember that the person was much more knowledgeable about Sudoku than myself, so I switched to using DIFC. As far as I can tell, they're the same thing.

ronk: I like your suggestion; but, if I decide to keep the distinction, I'm going to modify your suggestion to single-digit DIC/DIFC so the <1> can't be mistaken for the value I'm tracking.

Carcul: I don't know if it's necessary to maintain the distinction. If the DIC/DIFC or an equivalent chain is provided, then the single-digit restriction should be obvious. I'll have to give it some more thought.

wapati: You are just one of many forum posters who request solutions without chains. That's perfectly acceptable. Since many techniques can be viewed as chains, I no longer respond to requests for solutions with this constraint. My question was simply a query on why some people object to all chains. You are just one such person ... and you're certainly allowed to feel as you do. I feel that sometimes chains are the best way to proceed in a puzzle. I prefer simple chains in this case.