I am trying to apply a substitution rule to a Lebesgue integral and get very strange results.

Substitution rule (an instantiation of Fremlin, D.H. (2010), Measure Theory, Volume 2, Theorem 263D): Let $\phi: \mathbb{R} \to \mathbb{R}$ be an injective function with derivative $\phi': \mathbb{R} \to \mathbb{R}$. Let $f: \mathbb{R} \to \mathbb{R}$ be measurable and $\mu: \Sigma_{\mathbb{R}} \to \mathbb{R}$ a measure ($\Sigma_{\mathbb{R}}$ is a $\sigma$-algebra on $\mathbb{R}$). Then, $$ \int_{x \in \phi(\mathbb{R})} f(x) d\mu = \int_{x \in \mathbb{R}} |\phi'(x)|*f(\phi(x)) d\mu $$

I instantiate $\mu$ with the dirac measure $\delta(S) := [0 \in S]$, $f$ with the measurable function $f(x) := [x = 0]$ and $\phi$ with the function $\phi(x) := x+1$. Of course, $\phi'(x) = 1$.

Using this instantiation, I get \begin{align*} 1 &= f(0) \\ &= \int_{x \in \mathbb{R}} f(x) d\delta && \text{property of the dirac measure} \\ &= \int_{x \in \phi(\mathbb{R})} f(x) d\delta && \mathbb{R} = \phi(\mathbb{R}) \\ &= \int_{x \in \mathbb{R}} |\phi'(x)|*f(\phi(x)) d\delta && \text{substitution} \\ &= \int_{x \in \mathbb{R}} 1*f(x+1) d\delta \\ &= f(0+1) && \text{property of the dirac measure} \\ &= 0 \end{align*} What did I do wrong?


What you wrote doesn't match Theorem 263D in my copy of Fremlin. I think you missed an important sentence at the beginning of Section 263.

Throughout this section, as in the rest of the chapter, $\mu$ will denote Lebesgue measure on $\mathbb R^r$.


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.