1) With the 2x2x2, we normally wish to consider the puzzle solved ifSince we need to deal with whole cube rotations, I will use lower case
each face is all of one color. That is, whole cube rotations are
to be considered equivalent. With the Singmaster fixed face
center view of the 3x3x3, the issue of whole cube rotations does
not arise. But with the 2x2x2 we would normally consider
(for example) RL' equivalent to I. The most common way to
accomplish this type of equivalence is to fix one of the corners.
If we fix one of the corners of the 1x1x1, then we have a most
remarkable puzzle. There is only one state, nothing can ever
move, and the puzzle is always solved.
2) A second way to model the 2x2x2 such that whole cube rotations are
considered to be equivalent is to consider the set of states to be
the set of cosets of C, that is, the set of all xC.
If we take this approach with the 1x1x1, then there is only one
coset, namely iC (or just C, if you prefer). The cube can rotate,
but all 24 states are considered to be equivalent and the puzzle
is always solved.
3) Finally, if you model the 2x2x2 in such a way that whole cube
rotations are considered to be distinct, then you are really
modelling the corners of the 3x3x3. Indeed, a naive program
that simply modelled the permutations of the 2x2x2 facelets would
in fact unwittingly be modelling the corners of the 3x3x3.
If you take the same approach of modelling the permutations of the
1x1x1 facelets, then you in effect are considering whole
cube rotations to be distinct. You have a very easy problem,
but the problem is not totally trivial as it is with approach #1
or approach #2. The rest of this note will therefore consider the
problem of the 1x1x1 cube where whole cube rotations are considered
to be distinct.
Qturns OnlyThere are some additional problems we can look at. For an example, an
Distance Conjugacy Positions
from Classes
Start
0 1 1 {i}
1 1 6 {t,t',r,r',f,f'}
2 2 11 {tt,rr,ff},{tr,tr',tf,tf',t'r,t'r',t'f,t'f'}
3 1 6 {ttf,ttf',ffr,ffr',rrt,rrt'}
--- ---- ----
Total 5 24
Qturns Plus Hturns
Distance Conjugacy Positions
from Classes
Start
0 1 1 {i}
1 2 9 {t,t',r,r',f,f'},{t2,r2,f2}
2 2 14 {tr,tr',tf,tf',t'r,t'r',t'f,t'f'},
{t2f,t2f',f2r,f2r',r2t,r2t'}
--- ---- ----
Total 5 24
1) With the 2x2x2, we normally wish to consider the puzzle solved ifSince we need to deal with whole cube rotations, I will use lower case
each face is all of one color. That is, whole cube rotations are
to be considered equivalent. With the Singmaster fixed face
center view of the 3x3x3, the issue of whole cube rotations does
not arise. But with the 2x2x2 we would normally consider
(for example) RL' equivalent to I. The most common way to
accomplish this type of equivalence is to fix one of the corners.
If we fix one of the corners of the 1x1x1, then we have a most
remarkable puzzle. There is only one state, nothing can ever
move, and the puzzle is always solved.
2) A second way to model the 2x2x2 such that whole cube rotations are
considered to be equivalent is to consider the set of states to be
the set of cosets of C, that is, the set of all xC.
If we take this approach with the 1x1x1, then there is only one
coset, namely iC (or just C, if you prefer). The cube can rotate,
but all 24 states are considered to be equivalent and the puzzle
is always solved.
3) Finally, if you model the 2x2x2 in such a way that whole cube
rotations are considered to be distinct, then you are really
modelling the corners of the 3x3x3. Indeed, a naive program
that simply modelled the permutations of the 2x2x2 facelets would
in fact unwittingly be modelling the corners of the 3x3x3.
If you take the same approach of modelling the permutations of the
1x1x1 facelets, then you in effect are considering whole
cube rotations to be distinct. You have a very easy problem,
but the problem is not totally trivial as it is with approach #1
or approach #2. The rest of this note will therefore consider the
problem of the 1x1x1 cube where whole cube rotations are considered
to be distinct.
Qturns OnlyThere are some additional problems we can look at. For an example, an
Distance Conjugacy Positions
from Classes
Start
0 1 1 {i}
1 1 6 {t,t',r,r',f,f'}
2 2 11 {tt,rr,ff},{tr,tr',tf,tf',t'r,t'r',t'f,t'f'}
3 1 6 {ttf,ttf',ffr,ffr',rrt,rrt'}
--- ---- ----
Total 5 24
Qturns Plus Hturns
Distance Conjugacy Positions
from Classes
Start
0 1 1 {i}
1 2 9 {t,t',r,r',f,f'},{t2,r2,f2}
2 2 14 {tr,tr',tf,tf',t'r,t'r',t'f,t'f'},
{t2f,t2f',f2r,f2r',r2t,r2t'}
--- ---- ----
Total 5 24
原帖由 bardy 于 2009-5-16 12:13 发表
呵呵,我是想找一下详细的研究。以科学的态度
我看很多人可以学习一下,怎样把一件事情说得清楚明白
科学要怎么做,学问要怎么做。该详细的时候还是需要详细的
原帖由 CTS 于 2009-5-16 13:38 发表
...
只不過一個笑話
原帖由 Vicki 于 2009-5-16 11:08 发表
其实一阶严格来说不能算是魔方~
乌木还是大烟头说过:有相交的旋转面才能称作魔方~
所以最简单的魔方是1x2x2~
http://bbs.mf8-china.com/viewthread.php?tid=28193
欢迎光临 魔方吧·中文魔方俱乐部 (http://bbs.mf8-china.com/) | Powered by Discuz! X2 |