Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Every night Joe goes to the casino and takes with him an amount of money in dollars, X, that is distributed according to the pdf:

f(x) = Ax^2 for 0 < x < 10

where A is a constant that you need to determine. He returns with an amount of money Y that is uniformly distributed between 0 and 3X. Find

  • E[Y] w/out finding the distribution for Y (i.e. using E[E(Y|X)])
  • The pdf for Y
  • E[Y] using the distribution for Y. Compare with your result of part a.
  • The prob he returns w/ at least 5 dollars
  • If he returns w/ 5 dollars, what is the prob he left with at least 7 dollars
  • If he return w/ 5 dollars, what is the expected value of the amount of money with which he left?

I am having a hard time finding the joint pdf for this situation. I know f(x) = 3x^2/1000 for 0 < x < 10, but I keep getting 1/3X for the distribution of Y, which doesn't seem right. Any hints on how to form the joint pdf here?

share|cite|improve this question

Given that $X$ has value $x_0, 0 < x_0 < 10$, $Y$ is (conditionally) uniformly distributed on $[0, 3x_0]$ and hence $E[Y\mid X=x_0] = 3x_0/2$. (If you can't do this last step by inspection, work out the value of $\int_0^{3x_0} y \cdot \frac{1}{3x_0} \,\mathrm dy$). Since $E[Y\mid X=x_0]$ depends on the value of $X$, it can be regarded as a random variable_ that is a function of the random variable $X$). We denote this random variable by $E[Y\mid X]$ and note that it happens to be the function $3X/2$ in this instance. Can you now find $E[Y] = E[E[Y\mid X]] = E[3X/2]$ without needing to find the pdf of $Y$?

For the pdf of $Y$, note that $$f_{Y\mid X}(y\mid x) = \begin{cases}\frac{1}{3x}, &0 < y < 3x,\\\quad\\0, &\text{otherwise,}\end{cases}$$ and so, $$f_{X,Y}(x,y) = f_{Y\mid X}(y\mid x)f_X(x) = \begin{cases}\frac{1}{3x}\cdot\frac{3x^2}{1000}, &0 < y < 3x, 0 < x < 10,\\\quad\\0, &\text{otherwise.}\end{cases}$$ You can find the pdf of $Y$ from this. Remember to sketch the region where the joint pdf is nonzero; it will help you set the limits of integration correctly.

share|cite|improve this answer
I'm attempting to solve for E[Y] using the formula E[Y] = integral(E[Y|X]f(x)dx) from 0 to 10. Would this give the proper E[Y]? I got 45/4 for example. – sixglazed May 23 '13 at 1:21
@sixglazed Yes, $E[Y] = \frac{45}{4}$. – Dilip Sarwate May 23 '13 at 2:49

The conditional density of Y given X=x is 1/(3x). You need to find the unconditional distribution by integrating over f(x).

share|cite|improve this answer
Could you guide me in how to set up this integral? What would f(x) be in this case and what bounds would you use? – sixglazed May 23 '13 at 0:03
f(x) would be the PDF of x and the limits would be 0 to 10. – response May 23 '13 at 0:12
Isn't the integral of 3x^2/1000 from 0 to 10 just 1? – sixglazed May 23 '13 at 0:16
You integrate f(x) 1/(3x). – response May 23 '13 at 0:56

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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