There are several variants of dominated convergence theorem.

The standard one requires $f_n \to f$ a.e. and $|f_n|\leq g$ a.e. where $g$ is integrable. It can be weakened to only convergent in measure, if we impose $\sigma$-finiteness to the measure, c.f. Generalisation of Dominated Convergence Theorem.

However, we know if $g$ is integrable, $N(g) = \{g\neq 0\}$ has a $\sigma$- finite measure. Does it imply that converging in measure is alway sufficient, if given a dominating integrable function $g$.

If all statements made above are correct, I want to ask the following question, why most of textbooks only introduce the one with strong assumptions rather than the more general one?


Yes, convergence in measure plus integrability of $\sup_n|f_n|$ give convergence in $\mathbb L^1$, as dicussed here.

Maybe it's not stated in this way in the textbooks because in a lot (but not all) examples of application of this theorem, we have almost everywhere convergence.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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