# How to calculate the derivative of a random variable to one of the parameters?

So, I have a function: $f(x) = a\cdot x + b\cdot x + \varepsilon$, where $\varepsilon \sim \mathcal{N}(0, a+2b)$

I want to derive this function to one of the parameters (for example $a$) when performing a sensitivity analysis, but I get stuck with the epsilon part. Somebody has an idea?

-
For some basic information about writing math at this site see e.g. here, here, here and here. – Julian Kuelshammer Jan 22 '13 at 10:20
What is $N(0,a+2b)$? – ArthurStuart Jan 22 '13 at 12:08