Random variables and Linearity

350 Views Asked by At

I have an equation $Y = 5 + 3\times X$ and I assume that $X$ is a random variable taking values from a uniform distribution. Can I consider that also $Y$ is a random variable which takes values from a uniform distribution also but in different space?

Thank you in advance

2

There are 2 best solutions below

4
On

Not an expert, but I think $Y$ will be distributed uniformly as well.

Right attempt

Thank you for pointing out the error Did!

The uniform (continuous) distribution is determined by its cumulative distrubition function.

$$P(X \leq x) = F_X(x)= \begin{cases} 0 &\text{if } x < a\\ \frac{x-a}{b-a} &\text{if } x \in [a,b[ \\ 1 &\text{if } x \geq b\end{cases}$$

Looking at $F_Y(y)$:

$$F_Y(y) = P(Y\leq y) = P(5+3X \leq y) = P\left(X \leq \frac{y-5}{3}\right)$$

You can find the boundaries where this makes sense using: $$\frac{y-5}{3} \geq a \qquad \frac{y-5}{3} < b$$

Which concludes towards: $$P(Y \leq y) = F_Y(y)= \begin{cases} 0 &\text{if } y < 3a+5\\ \frac{\frac{y-5}{3}-a}{b-a} &\text{if } x \in [3a+5,3b+5[ \\ 1 &\text{if } x \geq 3b+5\end{cases}$$

Wrong attempt

($X$ will not be discrete, the overal idea was fine though)

If $X \stackrel{d}{=} U(a,b)$ then $$P(X =x) = \begin{cases} \frac{1}{b-a} \quad &\text{ if } x \in [a,b] \\ 0 &\text{ if not}\end{cases}$$ Looking at $P(Y=y)$ which determines the distribution you get: $$P(Y=y) = P(5+3X = y) = P\left(X = \frac{y-5}{3}\right)$$ Which means $Y$ will also be distributed uniformly.

0
On

If X is uniform on (0,1) then Y=5+3X is uniform on (5,8). More generally, for every nonzero b, Z=a+bX is uniform on (a,a+b).