MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $X=\{0,1\}^{\mathbb{N}}$ and $\xi_n$ be the partition of $X$ defined by the equivalence relation $x \sim_n x' \Leftrightarrow (x_{n}, x_{n+1}, \ldots) = (x_{n}', x_{n+1}', \ldots)$. The sequence of partitions $(\xi_n)$ is decreasing and we introduce the intersection partition $\theta= \cap \xi_n$. I'm loooking for a group $G$ of transformations of $X$ such that $\theta$ is the partition into the orbits of $G$.

Maybe my question is somewhat unprecise, I am rather new in ergodic theory. Any comments are welcomed.

share|cite|improve this question
Isn't it true that $\xi_n\subset \xi_{n+1}$? Why do you call this sequence decreasing? – Mark Sapir May 4 '12 at 21:40
Provided you mean to union up the equivalence relations, this equivalence relation is often called $E_0$ and can be easily realized as the orbit equivalence relation of the coordinate-wise action of $(\mathbb{Z}/2\mathbb{Z})^{< \omega}$ on $(\mathbb{Z}/2\mathbb{Z})^\omega$, or as the equivalence relation generated by addition by one in the $2$-adic integers (except you need to deal with the eventually constant sequences separately). – Clinton Conley May 4 '12 at 22:10
The ordering between $\xi_n$ and $\xi_{n+1}$ is not the inclusion. Any element of $\xi_{n}$ is a subset of an element of $\xi_{n+1}$, that is, $\xi_n$ is finer than $\xi_{n+1}$, or $\xi_{n+1}$ is coarser than $\xi_n$. It is sensible to denote $\xi_n \geq \xi_{n+1}$ because it is more naturel to say that the smallest (coarsest) partition is the trivial partition $\{\varnothing, X\}$ and the biggest (finest) partition is ${\cal P}(X)$. – Stéphane Laurent May 4 '12 at 22:17
@Stephane: Yes, I figured that out. That means I understood your question correctly. – Mark Sapir May 4 '12 at 22:19
@Clinton: That was my first choice too, but dealing with eventually constant sequences is not easy. – Mark Sapir May 4 '12 at 22:20
up vote 1 down vote accepted

Consider the group $G_n$ of all permutations of strings $\{0,1\}^{\{1,2,...,n\}}$ of length $n$. That group acts on the set of all strings $\{0,1\}^{\mathbb N}$ in the natural way (it changes the $n$-prefix only). Now clearly $G_n\subset G_{n+1}$ for every $n$. Let $G=\cup G_n$. It is a (locally finite) group which acts on $\{0,1\}^{\mathbb{N}}$ and the orbit equivalence relation is your $\theta$ (if I understood the question correctly): two strings are in the same orbit iff they coincide almost everywhere.

share|cite|improve this answer
This does not seem quite right; for example $01^\infty$ is not in the same orbit as $1^\infty$ under the action you propose. The orbit equivalence relation generated by your action is strictly finer than eventual agreement. – Clinton Conley May 4 '12 at 22:12
Yes, it is. $G_n$ is not the permutation group of indices, it is the permutation group of the set of strings of length $n$. – Mark Sapir May 4 '12 at 22:15
So $|G_n|=(2^n)!$. – Mark Sapir May 4 '12 at 22:16
Oops, sorry for misreading that. You're absolutely right. – Clinton Conley May 4 '12 at 22:17
@Mark: what do you mean by two strings coinciding almost everywhere ? – Stéphane Laurent May 4 '12 at 22:55

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.