Let $G$ be compact (and Hausdorff) group, $\mu$ be Haar measure on $G$. Is it always true that $(G,\mu)$ is a standard probability space (Lebesgue-Rokhlin space)? It is so if (a priori not iff) the topology of $G$ is metrizable, but this is probably not necessary.
$\begingroup$
$\endgroup$
2
-
4$\begingroup$ Well, some silly counter-examples: A finite group. Or $\{0,1\}^I$ for some very, very large set $I$. $\endgroup$– Matthew DawsCommented May 11, 2011 at 19:10
-
$\begingroup$ well, finite does not count, I had to mention it the second example is ok, thank you $\endgroup$– Fedor PetrovCommented May 11, 2011 at 19:54
Add a comment
|