1. Density function of random variable transformation

Given $X$ a real valued random variable with a density function $f_X$. Let $g$ be a continuous function.

Does this imply, that the random variable $g(X)$ has a density function? Or do I need more assumptions, for example that $g$ has an inverse function?

2. Density function of marginal distribution

Let $\lambda(dx,dy)$ be a Radon measure with a density function and $D$ is compact. Does this imply, that the marginal distribution $\lambda(dx)=\int_D \lambda(dx,dy)dy$ has a density function too?


If $g$ is a cosntant then $g(X)$ does not have a density. Even if you assume that $g$ has an inverse $g(X)$ need not have a density. For example if $X$ has uniform distribution and $g$ is a striclty increasing singular function then $g(X)$ does not have a density.

For the second part the answer is YES. $m(A)=0$ implies $m_2(A\times \mathbb R)=0$ where $m_2$ is two dimensional Lebesgue measure. This implies $\lambda (A \times \mathbb R)=0$ so the first marginal of $\lambda$ has an absolutely continuous distribution.

  • $\begingroup$ Thank you, that helps! $\endgroup$ – freeba Aug 12 at 13:47

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.