Perfect matching in the n-unit-cube, Is hyperplane statement wrong?











up vote
0
down vote

favorite












I was thinking about perfect matchings in the graph of the unit-cube of dimension $n$: $Q_n = [0,1]^n$. ($0$-$1$-strings of length n are vertices. Two of such are connected by an edge iff. they differ by only 1 bit (i.e. Hamming distance is equal to 1)).



Suppose you're given an arbitrary perfect matching $M$ in $Q_n$. Is it possible to find a facet $F$ of $Q_n$ such that all matching edges $e = (v,w) in M$ that cover a vertex in $F$ lie in $F$? Stated differently, if $v in F$ and $(v,w) in M$ for some vertex $w$, does that imply that $w in F$? This statement should be equivalent to finding a separating hyperplane, that separates $Q_n$ into two $(n-1)$-cubes, such that no matching edge crosses the hyperplane.



What do you think, is this true or false? Obviously its false for $n=1$ but I think it should hold for all $n geq 2$, but can't prove it!



Any advice on how to prove it or a counterexample is greatly appreciated!
Thanks!










share|cite|improve this question


















  • 1




    By “facet” do you mean an induced $Q_{n-1}$-subgraph?
    – Santana Afton
    Nov 14 at 19:39










  • Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
    – Doc
    Nov 14 at 19:53










  • Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
    – Santana Afton
    Nov 14 at 20:10










  • Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
    – Doc
    Nov 14 at 20:20






  • 1




    Let me know if this is any better.
    – Santana Afton
    Nov 14 at 21:13















up vote
0
down vote

favorite












I was thinking about perfect matchings in the graph of the unit-cube of dimension $n$: $Q_n = [0,1]^n$. ($0$-$1$-strings of length n are vertices. Two of such are connected by an edge iff. they differ by only 1 bit (i.e. Hamming distance is equal to 1)).



Suppose you're given an arbitrary perfect matching $M$ in $Q_n$. Is it possible to find a facet $F$ of $Q_n$ such that all matching edges $e = (v,w) in M$ that cover a vertex in $F$ lie in $F$? Stated differently, if $v in F$ and $(v,w) in M$ for some vertex $w$, does that imply that $w in F$? This statement should be equivalent to finding a separating hyperplane, that separates $Q_n$ into two $(n-1)$-cubes, such that no matching edge crosses the hyperplane.



What do you think, is this true or false? Obviously its false for $n=1$ but I think it should hold for all $n geq 2$, but can't prove it!



Any advice on how to prove it or a counterexample is greatly appreciated!
Thanks!










share|cite|improve this question


















  • 1




    By “facet” do you mean an induced $Q_{n-1}$-subgraph?
    – Santana Afton
    Nov 14 at 19:39










  • Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
    – Doc
    Nov 14 at 19:53










  • Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
    – Santana Afton
    Nov 14 at 20:10










  • Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
    – Doc
    Nov 14 at 20:20






  • 1




    Let me know if this is any better.
    – Santana Afton
    Nov 14 at 21:13













up vote
0
down vote

favorite









up vote
0
down vote

favorite











I was thinking about perfect matchings in the graph of the unit-cube of dimension $n$: $Q_n = [0,1]^n$. ($0$-$1$-strings of length n are vertices. Two of such are connected by an edge iff. they differ by only 1 bit (i.e. Hamming distance is equal to 1)).



Suppose you're given an arbitrary perfect matching $M$ in $Q_n$. Is it possible to find a facet $F$ of $Q_n$ such that all matching edges $e = (v,w) in M$ that cover a vertex in $F$ lie in $F$? Stated differently, if $v in F$ and $(v,w) in M$ for some vertex $w$, does that imply that $w in F$? This statement should be equivalent to finding a separating hyperplane, that separates $Q_n$ into two $(n-1)$-cubes, such that no matching edge crosses the hyperplane.



What do you think, is this true or false? Obviously its false for $n=1$ but I think it should hold for all $n geq 2$, but can't prove it!



Any advice on how to prove it or a counterexample is greatly appreciated!
Thanks!










share|cite|improve this question













I was thinking about perfect matchings in the graph of the unit-cube of dimension $n$: $Q_n = [0,1]^n$. ($0$-$1$-strings of length n are vertices. Two of such are connected by an edge iff. they differ by only 1 bit (i.e. Hamming distance is equal to 1)).



Suppose you're given an arbitrary perfect matching $M$ in $Q_n$. Is it possible to find a facet $F$ of $Q_n$ such that all matching edges $e = (v,w) in M$ that cover a vertex in $F$ lie in $F$? Stated differently, if $v in F$ and $(v,w) in M$ for some vertex $w$, does that imply that $w in F$? This statement should be equivalent to finding a separating hyperplane, that separates $Q_n$ into two $(n-1)$-cubes, such that no matching edge crosses the hyperplane.



What do you think, is this true or false? Obviously its false for $n=1$ but I think it should hold for all $n geq 2$, but can't prove it!



Any advice on how to prove it or a counterexample is greatly appreciated!
Thanks!







combinatorics graph-theory matching-theory bit-strings






share|cite|improve this question













share|cite|improve this question











share|cite|improve this question




share|cite|improve this question










asked Nov 14 at 19:38









Doc

367113




367113








  • 1




    By “facet” do you mean an induced $Q_{n-1}$-subgraph?
    – Santana Afton
    Nov 14 at 19:39










  • Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
    – Doc
    Nov 14 at 19:53










  • Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
    – Santana Afton
    Nov 14 at 20:10










  • Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
    – Doc
    Nov 14 at 20:20






  • 1




    Let me know if this is any better.
    – Santana Afton
    Nov 14 at 21:13














  • 1




    By “facet” do you mean an induced $Q_{n-1}$-subgraph?
    – Santana Afton
    Nov 14 at 19:39










  • Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
    – Doc
    Nov 14 at 19:53










  • Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
    – Santana Afton
    Nov 14 at 20:10










  • Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
    – Doc
    Nov 14 at 20:20






  • 1




    Let me know if this is any better.
    – Santana Afton
    Nov 14 at 21:13








1




1




By “facet” do you mean an induced $Q_{n-1}$-subgraph?
– Santana Afton
Nov 14 at 19:39




By “facet” do you mean an induced $Q_{n-1}$-subgraph?
– Santana Afton
Nov 14 at 19:39












Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
– Doc
Nov 14 at 19:53




Yes. With facet I mean an $n-1$ dimensional face of $Q_n$ viewed as a polytope. Each facet has one coordinate fixed to $0$ (or $1$) and all others are still in the interval $[0,1]$.
– Doc
Nov 14 at 19:53












Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
– Santana Afton
Nov 14 at 20:10




Is this a counter example? I’m not sure I can find a sub-cube that only fully contains matching edges. I apologize in advance for the quality — I’m not at my desk at the moment.
– Santana Afton
Nov 14 at 20:10












Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
– Doc
Nov 14 at 20:20




Thanks for your effort, but to be quite honest, I am not sure whether I understand your drawing correctly.
– Doc
Nov 14 at 20:20




1




1




Let me know if this is any better.
– Santana Afton
Nov 14 at 21:13




Let me know if this is any better.
– Santana Afton
Nov 14 at 21:13










1 Answer
1






active

oldest

votes

















up vote
3
down vote



accepted










Here is a perfect matching in $Q_4$ that does not have this property.



enter image description here






share|cite|improve this answer





















  • I think we came up with very nearly the same matching!
    – Santana Afton
    Nov 14 at 21:14






  • 1




    It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
    – Misha Lavrov
    Nov 14 at 21:34










  • Thanks, I was pretty sure that this holds!
    – Doc
    Nov 15 at 7:31











Your Answer





StackExchange.ifUsing("editor", function () {
return StackExchange.using("mathjaxEditing", function () {
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
});
});
}, "mathjax-editing");

StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "69"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});

function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});


}
});














 

draft saved


draft discarded


















StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2998733%2fperfect-matching-in-the-n-unit-cube-is-hyperplane-statement-wrong%23new-answer', 'question_page');
}
);

Post as a guest















Required, but never shown

























1 Answer
1






active

oldest

votes








1 Answer
1






active

oldest

votes









active

oldest

votes






active

oldest

votes








up vote
3
down vote



accepted










Here is a perfect matching in $Q_4$ that does not have this property.



enter image description here






share|cite|improve this answer





















  • I think we came up with very nearly the same matching!
    – Santana Afton
    Nov 14 at 21:14






  • 1




    It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
    – Misha Lavrov
    Nov 14 at 21:34










  • Thanks, I was pretty sure that this holds!
    – Doc
    Nov 15 at 7:31















up vote
3
down vote



accepted










Here is a perfect matching in $Q_4$ that does not have this property.



enter image description here






share|cite|improve this answer





















  • I think we came up with very nearly the same matching!
    – Santana Afton
    Nov 14 at 21:14






  • 1




    It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
    – Misha Lavrov
    Nov 14 at 21:34










  • Thanks, I was pretty sure that this holds!
    – Doc
    Nov 15 at 7:31













up vote
3
down vote



accepted







up vote
3
down vote



accepted






Here is a perfect matching in $Q_4$ that does not have this property.



enter image description here






share|cite|improve this answer












Here is a perfect matching in $Q_4$ that does not have this property.



enter image description here







share|cite|improve this answer












share|cite|improve this answer



share|cite|improve this answer










answered Nov 14 at 21:07









Misha Lavrov

41.7k555101




41.7k555101












  • I think we came up with very nearly the same matching!
    – Santana Afton
    Nov 14 at 21:14






  • 1




    It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
    – Misha Lavrov
    Nov 14 at 21:34










  • Thanks, I was pretty sure that this holds!
    – Doc
    Nov 15 at 7:31


















  • I think we came up with very nearly the same matching!
    – Santana Afton
    Nov 14 at 21:14






  • 1




    It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
    – Misha Lavrov
    Nov 14 at 21:34










  • Thanks, I was pretty sure that this holds!
    – Doc
    Nov 15 at 7:31
















I think we came up with very nearly the same matching!
– Santana Afton
Nov 14 at 21:14




I think we came up with very nearly the same matching!
– Santana Afton
Nov 14 at 21:14




1




1




It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
– Misha Lavrov
Nov 14 at 21:34




It's quite possible that there is only one construction up to symmetry here. We have 4 directions, and in each one of them we want at least one (and therefore at least two) edges.
– Misha Lavrov
Nov 14 at 21:34












Thanks, I was pretty sure that this holds!
– Doc
Nov 15 at 7:31




Thanks, I was pretty sure that this holds!
– Doc
Nov 15 at 7:31


















 

draft saved


draft discarded



















































 


draft saved


draft discarded














StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2998733%2fperfect-matching-in-the-n-unit-cube-is-hyperplane-statement-wrong%23new-answer', 'question_page');
}
);

Post as a guest















Required, but never shown





















































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown

































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown







Popular posts from this blog

How to change which sound is reproduced for terminal bell?

Can I use Tabulator js library in my java Spring + Thymeleaf project?

Title Spacing in Bjornstrup Chapter, Removing Chapter Number From Contents