Skip to main content
deleted 422 characters in body
Source Link
Will Sawin
  • 148.4k
  • 9
  • 324
  • 563

I came up with a solution which was incorrect. However, one of my partial results might be interesting (or might be trivial)Yes.

First let's sayLet me use the start ofreduction to labelings suggested by Rohit. Let's call a $2$two-bushy tree where $T$$\sigma$ is the noderoot a $\sigma$$2$-branchy tree.

Result: LetFor any continuous function $v_1$ and$L$ from $v_2$ be two nodes. There$4^{<\omega}$ to $\{0,1\}$, there is a $2$-bushy tree $T_1$ starting at $v_1$ andeither a $2$three-bushybranchy tree $T_2$ starting at $v_2$ such that either$T$ where $f$ is constant$L$ restricted to $T_1$, or $f$$[T]$ is the constant restricted tofunction $T_2$,$0$ or the image ofa two-branchy tree $f$$T$ where $L$ restricted to $T_1$$[T]$ is disjoint from the image of $f$ restricted toconstant function $T_2$$1$.

I thought that This follows by applyingbackwards induction on the last legnodes, starting from nodes where $L$ is constant. If a node has all four children the base of one of these two types of trees, then either three have the trichotomy many times you could produce an injective functionfirst type of tree or two have the second, but this turned out not to beso it has one of the casetwo types of threes.

To prove this, note:

Lemma: Let $T$ beAs a corollary there is either a $a$$2$-bushybranchy tree starting at a node $v$. Let$T$ such that $Z$ be a closed set of$L$ restricted to $2^{ <\omega}$. Suppose that$T$ is $0$, for everyor a $b$$2$-bushy subtreebranchy tree $T'$ of$T$ such that $L$ restricted to $T$ starting atis $v$$1$. Without loss of generality, assume it is $f([T']) \cap Z \neq \varnothing$$0$. Then let $k$ be the maximum number such that there is ana $a-b+1$$2$-bushy subtree $T^*$ ofbranchy tree $T$ suchwhere $f$ restricted to $[T]$ lands on functions that begin with $f(T^*) \subseteq Z$$k$ zeros. By assumption $k\geq 1$.

Proof of lemma: LetIf $T^*$ be the subtree of$k=\infty$ we win $T$ consisting of all vertices that are not the start of an- there is a tree where $b$$f$ is constant.

So assume $k$ is finite. Suppose there exists a $2$-bushy subtree whose branches are all sent tobranchy tree $2^{ \omega}-Z$ by$T'$ such that $f$ restricted to $[T']$ lands on functions that do not begin with $k$ zeros. Then every node ofwe win. Indeed the first $T^*$ has at least$k$ values of $a+b-1$ successors in$f(x_0x_1\dots)$ are determined by $T^*$$x_0x_1\dots x_{N-1}$ for some $N$. Then choose a two-branchy tree of sequences $x_0x_1\dots$ such that if it had any fewer, it would have $b$ sucessors, each$x_i$ is the startleftmost of athe two options at $b$-bushy subtree whose branches are sent to$x_0\dots x_{i-1}$ then $2^{\omega}-Z$$x_{Ni}x_{N{i+1}} \dots x_{Ni+N-1}$ is in $T$, hence it would beand if $x_i$ is the startrightmost of athe two options then $b$-bushy subtree whose branches are sent to$x_{Ni}x_{N{i+1}} \dots x_{Ni+N-1}$ is in $2^{\omega}-Z$. By assumption,$T'$ $T^*$ contains(and we might have to switch left and right for $v$$x_0$). Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not inThen $Z$, in which it would be easy$f$ restricted to form a $b$-bushythat tree startingis injective, as we can reconstruct $x_i$ by looking at one of the vertices$Ni$ through $Ni+k-1$ digits of the branch whose branches do not intersect $Z$$f(x_0x_i\dots)$.

Now using the lemma, takeIf there does not exist such a subtree $T_1$ starting at $v_1$ and let $Z=f([T_1])$. Then either we can find a 2$2$-bushybranchy tree $T_2$ with $f([T_2])$ disjoint from $f([T_1])$, and we're done, or we can findthen there exists a $3$-bushybranchy tree $S_2$ with $f([S_2]) \subseteq Z$. Now pick a point in$T^*$ where $Z$. We can either find a$f$ restricted to $2$-bushy subtree$T^*$ always begins with $T_2$$k$ zeroes. Then consider the tree of $S_2$, withsequences $f([T_2])$ not containing$x$ that point, or find a 2-bushy subtree withare in $f([T_2])$ constant$T^*$ and equal to that point. In the second case, we are done. Inafter removing the first casedigit, remain in $f([T_2])$$T^*$. This is a closed subset of $f([T_1])$. Iterating, we get a 2$2$-bushy subtree $T_3$ starting at $v_1$ with $f([T_3])$ a proper closed subset of $f([T_1])$. We can do this infinitely many times and take the limitbranchy tree, and $f$ applied to the limit is containedeverything in the intersection, and then repeat the processit has $k+1$ zeroes. By transfinite inductionThis contradicts our assumption on / Zorn's Lemma, we eventually must reach a closed set containing no points, which is the empty set, a contradiction$k$. 

So we're donewe win.

I came up with a solution which was incorrect. However, one of my partial results might be interesting (or might be trivial).

First let's say the start of a $2$-bushy $T$ is the node $\sigma$.

Result: Let $v_1$ and $v_2$ be two nodes. There is a $2$-bushy tree $T_1$ starting at $v_1$ and a $2$-bushy tree $T_2$ starting at $v_2$ such that either $f$ is constant restricted to $T_1$, or $f$ is constant restricted to $T_2$, or the image of $f$ restricted to $T_1$ is disjoint from the image of $f$ restricted to $T_2$.

I thought that by applying the last leg of the trichotomy many times you could produce an injective function, but this turned out not to be the case.

To prove this, note:

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an $a-b+1$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof of lemma: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches are all sent to $2^{ \omega}-Z$ by $f$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $b$ sucessors, each the start of a $b$-bushy subtree whose branches are sent to $2^{\omega}-Z$, hence it would be the start of a $b$-bushy subtree whose branches are sent to $2^{\omega}-Z$. By assumption, $T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not in $Z$, in which it would be easy to form a $b$-bushy tree starting at one of the vertices of the branch whose branches do not intersect $Z$.

Now using the lemma, take a subtree $T_1$ starting at $v_1$ and let $Z=f([T_1])$. Then either we can find a 2-bushy tree $T_2$ with $f([T_2])$ disjoint from $f([T_1])$, and we're done, or we can find a $3$-bushy tree $S_2$ with $f([S_2]) \subseteq Z$. Now pick a point in $Z$. We can either find a $2$-bushy subtree $T_2$ of $S_2$, with $f([T_2])$ not containing that point, or find a 2-bushy subtree with $f([T_2])$ constant and equal to that point. In the second case, we are done. In the first case, $f([T_2])$ is a closed subset of $f([T_1])$. Iterating, we get a 2-bushy subtree $T_3$ starting at $v_1$ with $f([T_3])$ a proper closed subset of $f([T_1])$. We can do this infinitely many times and take the limit, and $f$ applied to the limit is contained in the intersection, and then repeat the process. By transfinite induction / Zorn's Lemma, we eventually must reach a closed set containing no points, which is the empty set, a contradiction. So we're done.

Yes.

Let me use the reduction to labelings suggested by Rohit. Let's call a two-bushy tree where $\sigma$ is the root a $2$-branchy tree.

For any continuous function $L$ from $4^{<\omega}$ to $\{0,1\}$, there is a either a three-branchy tree $T$ where $L$ restricted to $[T]$ is the constant function $0$ or a two-branchy tree $T$ where $L$ restricted to $[T]$ is the constant function $1$. This follows by backwards induction on the nodes, starting from nodes where $L$ is constant. If a node has all four children the base of one of these two types of trees, then either three have the first type of tree or two have the second, so it has one of the two types of threes.

As a corollary there is either a $2$-branchy tree $T$ such that $L$ restricted to $T$ is $0$, or a $2$-branchy tree $T$ such that $L$ restricted to $T$ is $1$. Without loss of generality, assume it is $0$. Then let $k$ be the maximum number such that there is a $2$-branchy tree $T$ where $f$ restricted to $[T]$ lands on functions that begin with $k$ zeros. By assumption $k\geq 1$.

If $k=\infty$ we win - there is a tree where $f$ is constant.

So assume $k$ is finite. Suppose there exists a $2$-branchy tree $T'$ such that $f$ restricted to $[T']$ lands on functions that do not begin with $k$ zeros. Then we win. Indeed the first $k$ values of $f(x_0x_1\dots)$ are determined by $x_0x_1\dots x_{N-1}$ for some $N$. Then choose a two-branchy tree of sequences $x_0x_1\dots$ such that if $x_i$ is the leftmost of the two options at $x_0\dots x_{i-1}$ then $x_{Ni}x_{N{i+1}} \dots x_{Ni+N-1}$ is in $T$, and if $x_i$ is the rightmost of the two options then $x_{Ni}x_{N{i+1}} \dots x_{Ni+N-1}$ is in $T'$ (and we might have to switch left and right for $x_0$). Then $f$ restricted to that tree is injective, as we can reconstruct $x_i$ by looking at the $Ni$ through $Ni+k-1$ digits of $f(x_0x_i\dots)$.

If there does not exist such a $2$-branchy tree, then there exists a $3$-branchy tree $T^*$ where $f$ restricted to $T^*$ always begins with $k$ zeroes. Then consider the tree of sequences $x$ that are in $T^*$ and that, after removing the first digit, remain in $T^*$. This is $2$-branchy tree, and $f$ applied to everything in it has $k+1$ zeroes. This contradicts our assumption on $k$. 

So we win.

Post Undeleted by Will Sawin
added 419 characters in body
Source Link
Will Sawin
  • 148.4k
  • 9
  • 324
  • 563

I think it is true. Moreover I think you can replace $2^\omega$ with any T1 topological space, and it's true for $n$-bushy treescame up with $4^\omega$ replaced by $(3n-2)^\omega$.

For a $2$-bushy treesolution which was incorrect. However, let's call the node $\sigma$ the startone of the tree.my partial results might be interesting (I'd call it the root, but youor might want all trees to be rooted at the empty sequence.trivial).

Suppose there are noFirst let's say the start of a $2$-bushy trees $T$ starting atis the root of $4^{\leq \omega}$ such that $f$ restricted tonode $[T]$ is one-one$\sigma$. Call a tree

Result: Let $T$$v_1$ and $k$-injective if any$v_2$ be two branches through $T$ that get mapped by $f$ to the same point agree for the first $k$ nodes. Then a limit ofThere is a sequence $T_k$ of $k$-injective $2$-bushy treestree $T_1$ starting at the root is$v_1$ and a $2$-bushy tree on which$T_2$ starting at $v_2$ such that either $f$ is injective.

Hence for someconstant restricted to $k$$T_1$, there are noor $k$-injective$f$ is constant restricted to $2$-bushy trees starting at$T_2$, or the root. Make aimage of $2$-bushy tree starting at the root that$f$ restricted to $T_1$ is arbitrary indisjoint from the first $k$ levels. There must be no wayimage of continuing this$f$ restricted to a $k$-injective tree$T_2$. Hence we have $2^k$ nodes such

I thought that, for every $2^k$-tuple by applying the last leg of $2$-bushy trees starting at those nodesthe trichotomy many times you could produce an injective function, a branch from two of them must mapbut this turned out not to be the same pointcase.

To prove this, note:

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an $a-b+a$$a-b+1$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof of lemma: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches are all lie insent to $2^{ \omega}-Z$ by $f$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $b$ sucessors, each the start of a $b$-bushy subtree whose branches lie inare sent to $2^{\omega}-Z$, hence it would be the start of a $b$-bushy subtree whose branches lie inare sent to $2^{\omega}-Z$. By assumption, $T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not in $Z$, in which it would be easy to form a $b$-bushy tree starting at one of the vertices of the branch whose branches do not intersect $Z$.

Now pickusing the lemma, take a nodesubtree $T_1$ starting at $v_1$ and let $Z_1$$Z=f([T_1])$. Then either we can find a 2-bushy tree $T_2$ with $f([T_2])$ disjoint from $f([T_1])$, and we're done, or we can find a $Z_2$ be two closed sets such$3$-bushy tree $S_2$ with $f([S_2]) \subseteq Z$. Now pick a point in $Z$. We can either find a $2$-bushy subtree $T_2$ of $S_2$, with $f([T_2])$ not containing that point, for everyor find a 2-bushy subtree with $2$$f([T_2])$ constant and equal to that point. In the second case, we are done. In the first case, $f([T_2])$ is a closed subset of $f([T_1])$. Iterating, we get a 2-bushy treesubtree $T$$T_3$ starting at that node, $f([T])$ intersects$v_1$ with $Z_1$$f([T_3])$ a proper closed subset of $f([T_1])$. We can do this infinitely many times and take the limit, and $Z_2$$f$ applied to the limit is contained in the intersection, and then repeat the process. By transfinite induction / Zorn's Lemma, we eventually must reach a closed set containing no points, which is the empty set, a contradiction. So we're done.

I think it is true. Moreover I think you can replace $2^\omega$ with any T1 topological space, and it's true for $n$-bushy trees with $4^\omega$ replaced by $(3n-2)^\omega$.

For a $2$-bushy tree, let's call the node $\sigma$ the start of the tree. (I'd call it the root, but you might want all trees to be rooted at the empty sequence.)

Suppose there are no $2$-bushy trees $T$ starting at the root of $4^{\leq \omega}$ such that $f$ restricted to $[T]$ is one-one. Call a tree $T$ $k$-injective if any two branches through $T$ that get mapped by $f$ to the same point agree for the first $k$ nodes. Then a limit of a sequence $T_k$ of $k$-injective $2$-bushy trees starting at the root is a $2$-bushy tree on which $f$ is injective.

Hence for some $k$, there are no $k$-injective $2$-bushy trees starting at the root. Make a $2$-bushy tree starting at the root that is arbitrary in the first $k$ levels. There must be no way of continuing this to a $k$-injective tree. Hence we have $2^k$ nodes such that, for every $2^k$-tuple of $2$-bushy trees starting at those nodes, a branch from two of them must map to the same point.

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an $a-b+a$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches all lie in $2^{ \omega}-Z$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $b$ sucessors, each the start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$, hence it would be the start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$. $T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not in $Z$, in which it would be easy to form a $b$-bushy tree starting at one of the vertices of the branch whose branches do not intersect $Z$.

Now pick a node and let $Z_1$ and $Z_2$ be two closed sets such that, for every $2$-bushy tree $T$ starting at that node, $f([T])$ intersects $Z_1$ and $Z_2$.

I came up with a solution which was incorrect. However, one of my partial results might be interesting (or might be trivial).

First let's say the start of a $2$-bushy $T$ is the node $\sigma$.

Result: Let $v_1$ and $v_2$ be two nodes. There is a $2$-bushy tree $T_1$ starting at $v_1$ and a $2$-bushy tree $T_2$ starting at $v_2$ such that either $f$ is constant restricted to $T_1$, or $f$ is constant restricted to $T_2$, or the image of $f$ restricted to $T_1$ is disjoint from the image of $f$ restricted to $T_2$.

I thought that by applying the last leg of the trichotomy many times you could produce an injective function, but this turned out not to be the case.

To prove this, note:

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an $a-b+1$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof of lemma: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches are all sent to $2^{ \omega}-Z$ by $f$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $b$ sucessors, each the start of a $b$-bushy subtree whose branches are sent to $2^{\omega}-Z$, hence it would be the start of a $b$-bushy subtree whose branches are sent to $2^{\omega}-Z$. By assumption, $T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not in $Z$, in which it would be easy to form a $b$-bushy tree starting at one of the vertices of the branch whose branches do not intersect $Z$.

Now using the lemma, take a subtree $T_1$ starting at $v_1$ and let $Z=f([T_1])$. Then either we can find a 2-bushy tree $T_2$ with $f([T_2])$ disjoint from $f([T_1])$, and we're done, or we can find a $3$-bushy tree $S_2$ with $f([S_2]) \subseteq Z$. Now pick a point in $Z$. We can either find a $2$-bushy subtree $T_2$ of $S_2$, with $f([T_2])$ not containing that point, or find a 2-bushy subtree with $f([T_2])$ constant and equal to that point. In the second case, we are done. In the first case, $f([T_2])$ is a closed subset of $f([T_1])$. Iterating, we get a 2-bushy subtree $T_3$ starting at $v_1$ with $f([T_3])$ a proper closed subset of $f([T_1])$. We can do this infinitely many times and take the limit, and $f$ applied to the limit is contained in the intersection, and then repeat the process. By transfinite induction / Zorn's Lemma, we eventually must reach a closed set containing no points, which is the empty set, a contradiction. So we're done.

added 1755 characters in body
Source Link
Will Sawin
  • 148.4k
  • 9
  • 324
  • 563

I'm not sure I understand all the definitions that are used here correctlythink it is true. But ifMoreover I dothink you can replace $2^\omega$ with any T1 topological space, then this works:and it's true for $n$-bushy trees with $4^\omega$ replaced by $(3n-2)^\omega$.

ChooseFor a bijection $4 \to 2^2$ given by functions$2$-bushy tree, let's call the node $a_0,a_1: 4 \to 2$$\sigma$ the start of the tree. (I'd call it the root, and define

$$f(x_1x_2x_3...)=a_0(x_1)a_1(x_1)a_0(x_3)a_1(x_3)a_0(x_5)a_1(x_5)....$$but you might want all trees to be rooted at the empty sequence.)

ThenSuppose there are no $2$-bushy trees $T$ starting at the root of $4^{\leq \omega}$ such that $f$ restricted to any$[T]$ is one-one. Call a tree that$T$ $k$-injective if any two branches through $T$ that get mapped by $f$ to the same point agree for the first $k$ nodes. Then a limit of a sequence $T_k$ of $k$-injective $2$-bushy trees starting at the root is a $2$-bushy tree on an even row,which $f$ cannotis injective.

Hence for some $k$, there are no $k$-injective $2$-bushy trees starting at the root. Make a $2$-bushy tree starting at the root that is arbitrary in the first $k$ levels. There must be oneno way of continuing this to a $k$-oneinjective tree. Hence we have $2^k$ nodes such that, and restrictedfor every $2^k$-tuple of $2$-bushy trees starting at those nodes, a branch from two of them must map to anythe same point.

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that branches on, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an odd row$a-b+a$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches all lie in $2^{ \omega}-Z$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $f$ cannot$b$ sucessors, each the start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$, hence it would be constantthe start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$. Since bushy trees branch on almost$T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any rowbranch not in $Z$ would have an open neighborhood of branches not in $Z$, there is noin which it would be easy to form a $T$ satisfying$b$-bushy tree starting at one of the conditionsvertices of the branch whose branches do not intersect $Z$. But

Now pick a node and let $f$ is certainly continuous$Z_1$ and $Z_2$ be two closed sets such that, for every $2$-bushy tree $T$ starting at that node, $f([T])$ intersects $Z_1$ and $Z_2$.

I'm not sure I understand all the definitions that are used here correctly. But if I do, then this works:

Choose a bijection $4 \to 2^2$ given by functions $a_0,a_1: 4 \to 2$, and define

$$f(x_1x_2x_3...)=a_0(x_1)a_1(x_1)a_0(x_3)a_1(x_3)a_0(x_5)a_1(x_5)....$$

Then restricted to any tree that branches on an even row, $f$ cannot be one-one, and restricted to any tree that branches on an odd row, $f$ cannot be constant. Since bushy trees branch on almost any row, there is no $T$ satisfying the conditions. But $f$ is certainly continuous.

I think it is true. Moreover I think you can replace $2^\omega$ with any T1 topological space, and it's true for $n$-bushy trees with $4^\omega$ replaced by $(3n-2)^\omega$.

For a $2$-bushy tree, let's call the node $\sigma$ the start of the tree. (I'd call it the root, but you might want all trees to be rooted at the empty sequence.)

Suppose there are no $2$-bushy trees $T$ starting at the root of $4^{\leq \omega}$ such that $f$ restricted to $[T]$ is one-one. Call a tree $T$ $k$-injective if any two branches through $T$ that get mapped by $f$ to the same point agree for the first $k$ nodes. Then a limit of a sequence $T_k$ of $k$-injective $2$-bushy trees starting at the root is a $2$-bushy tree on which $f$ is injective.

Hence for some $k$, there are no $k$-injective $2$-bushy trees starting at the root. Make a $2$-bushy tree starting at the root that is arbitrary in the first $k$ levels. There must be no way of continuing this to a $k$-injective tree. Hence we have $2^k$ nodes such that, for every $2^k$-tuple of $2$-bushy trees starting at those nodes, a branch from two of them must map to the same point.

Lemma: Let $T$ be a $a$-bushy tree starting at a node $v$. Let $Z$ be a closed set of $2^{ <\omega}$. Suppose that, for every $b$-bushy subtree $T'$ of $T$ starting at $v$, $f([T']) \cap Z \neq \varnothing$. Then there is an $a-b+a$-bushy subtree $T^*$ of $T$ such that $f(T^*) \subseteq Z$.

Proof: Let $T^*$ be the subtree of $T$ consisting of all vertices that are not the start of an $b$-bushy subtree whose branches all lie in $2^{ \omega}-Z$. Then every node of $T^*$ has at least $a+b-1$ successors in $T^*$ - if it had any fewer, it would have $b$ sucessors, each the start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$, hence it would be the start of a $b$-bushy subtree whose branches lie in $2^{\omega}-Z$. $T^*$ contains $v$. Finally $[T^*] \subseteq Z$ because any branch not in $Z$ would have an open neighborhood of branches not in $Z$, in which it would be easy to form a $b$-bushy tree starting at one of the vertices of the branch whose branches do not intersect $Z$.

Now pick a node and let $Z_1$ and $Z_2$ be two closed sets such that, for every $2$-bushy tree $T$ starting at that node, $f([T])$ intersects $Z_1$ and $Z_2$.

Post Deleted by Will Sawin
Source Link
Will Sawin
  • 148.4k
  • 9
  • 324
  • 563
Loading