I have a continuous random variable X with a known PDF. I want to find the distribution of f(X) where f(X) is a function of X (e.g. X^2+5X). How do I find the distribution of E(f(X)|X))?

I know how to approach this when X is discrete. In this case, I would simply calculate E(f(X)|X=t) by summing over all possible values of X=t and then replace t by X to get the distribution of E(f(X)|X).

  • 1
    $\begingroup$ It sounds like you need a formula for transforming a variable. The hits at this search ought to help: stats.stackexchange.com/search?q=jacobian+change+score%3A1. $\endgroup$ – whuber Mar 21 '19 at 16:20
  • $\begingroup$ What do you mean by E(f(X)|X)). Also, what do you mean by lowercase x as opposed to the uppercase X? $\endgroup$ – Martin Drozdik Mar 21 '19 at 16:24
  • $\begingroup$ lower case x was a typo and is fixed. By E(f(X)|X), I mean conditional expectation of f(X) given X. I think it should be f(X) but I just wanted to confirm. Then as @whuber mentioned, I will have to look at deriving distribution of function of a random variable. $\endgroup$ – exAres Mar 21 '19 at 16:27

for any random variable(discrete , continues and mixture) and Borel function, f

$E(f(X)|X)=f(X)$ so you just need to compute distribution of $f(X)$

so $E(X^2+X|X)=X^2+X$

note $E(f(X)|X)=f(X)$ is a random variable and $E(f(X)|X=t)=f(t)$ is a number

this property ,$E(f(X)|X)=f(X)$ come from defination of conditional expectation and note that $\sigma(f(X)) \subset \sigma(X)$


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.