# Solving an exponential distribution

In a simulation, I am trying to find the value of $d_i$ where:

$\displaystyle d_i \sim \frac{\epsilon_i}{\lambda_i}$ where $\epsilon_i$ is i.i.d. exponentially distributed with parameter = 1 and $i=1...n$.

Conditional on $\lambda_i$ the $d_i$ have an exponential distribution of $\lambda_i$. I know the value of $\lambda_i$ but I don't know how to find the value of $d_i$. What are the steps should I undertake to find $d_i$? How relevant is the $\epsilon_i$?

-

All you need to do is simulate $\epsilon_i$ and then divide by $\lambda_i$.
So for example if $U_i$ is uniformly distributed on $[0,1)$ then you can take $\epsilon_i = -\log_e (1-U_i)$ and $d_i = \dfrac{-\log_e (1-U_i)}{\lambda_i}$.
ok perfect. Thanks. So does the exponential distribution of $\lambda_i$ matter? – ChuckM Jul 12 '12 at 7:53