MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

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

It is true that, under some conditions, given a measure-preserving transformation $T$, we can always construct a $T$-invariant probability. I am wondering whether we can do a converse. See Parry's Topics in ergodic theory p14

Given a probability space $(X,\mathcal{B},\mu)$, can we always find a measure-preserving transformation $T:X \rightarrow X$ such that $\mu$ is $T$-invariant, except the identity?

share|cite|improve this question
The identity map. – Jon Bannon Mar 27 '11 at 20:07
You are right, I should exclude the identity. – Po C. Mar 27 '11 at 20:12
up vote 6 down vote accepted

For a counterexample to Jon's implicit rewording of your question, consider a purely atomic probability space in which the atoms all have different measure.

For a positive result, consider any purely non atomic probability space. It is measure isomorphic to $[0,1]$ with Lebesgue measure. I guess for non separable purely non atomic probability spaces you just need to apply Maharam's classification theorem.

share|cite|improve this answer
@Jon: In the category of measurable spaces and equivalence classes of non-singular maps every non-complete measurable space is equivalent to its completion. – Dmitri Pavlov Mar 27 '11 at 22:02
Thanks, Dmitri. I was being silly... clearly, the right notion here is 'isomorphism mod 0'. – Jon Bannon Mar 27 '11 at 22:51

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.