I was hoping to get my solution to part $\textbf{i}$ of this qual question regarding the Radon-Nikodym derivative checked for rigor and correctness. Then I was hoping to get advice on proceeding with part $\textbf{ii}$. Here is the question:

Let $ m$ be the Lebesgue measure and define a measure $ \mu$ on the Borel $\sigma$-algebra on $[0,1]$ by the formula $$\mu(X) = m(\{y\in [0,\pi]:\sin(y) \in X\}) .$$ $\textbf{i}$ Show that $\mu$ is absolutely continuous with respect to the Lesbesgue measure.

$\textbf{ii.}$ Find the Radon Nikodym derivative $\frac{d\mu}{dm}. $

Here is my solution:

$\textbf{i}$ Let $A\subset \mathbb{R} $ such that $m(A) = 0$. We have that $$ \mu(A) = m(\{y\in [0,\pi]:\sin(y) \in A\}) $$ and from monotonicity we have $$\mu(A) \leq m(A\cap [0,\pi]) \leq m(A) = 0$$ and therefore $\mu << m $.

$\textbf{ii}$ Since $m $ is a $\sigma$-finite positive measure and $\mu $ is a finite positive measure (notice that $\mu(\mathbb{R}) = \pi)$ and $\mu << m $ by the Radon-Nikodym theorem there exists a $m$-integrable non-negative function $f $ which is measurable with respect to the Borel $\sigma $-algebra and $$\mu(A) = \int_A f d\mu $$

Furthermore, if any other function $g $ satisfies above, then $f=g $ a.e.

Notice that the set $B=[0,1] $ has full measure with respect to $\mu $. I want to claim that $$ f= 2\sin^{-1}(x) \cdot \chi_{[0,1]}$$ is the Radon Nikodym derivative, but I'm not exactly sure how to show this.

  • 1
    $\begingroup$ its not true that $\mu(A) \le m(A)$ in general, so how do you justify the inequality? Eg $\mu((0,1))=\pi>1=m((0,1))$ $\endgroup$ Jul 14 at 5:10

Your inequality is not correct. Here are some hints:

Let $\mu_1(X)=m\{x \in [0,\frac {\pi} 2]: \sin y \in X\}$. Verify that the derivative of $\sin^{-1} x$, namely $\frac 1 {\sqrt {1-x^{2}}}$ is the RND of $\mu_1$ w.r.t. $m$. [ For this show that $\int_a^{b} \frac 1 {\sqrt {1-x^{2}}} dx=\mu_1 (a,b)$ for $a <b$].

Now consider $\mu_2(X)=m\{x \in [\frac {\pi} 2, \pi]: \sin y \in X\}$. Use the fact that $\sin (\pi -y) =\sin y$ to prove absolute continuity of $\mu_2$ and for writing down the RND of $\mu_2$.


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.