Simulation of sample size

[Jonathan KJ]

Hi guys, I was just hoping to get some advice with stata simulation

I'm playing around with a couple of variables, x1 - uniform distribution, (5,10), x2 - normal distribution (5/2.5) and e which is. normal dist (0/1)
So the question is trying to generate these, and put it into a single outcome variable where
y = 25 - 2(x1) + 0.95(x2) + e

So assuming I'm working with a sample size of 10,000

set obs 10000

for x1
gen uniform1 = runiform(5,10)
histogram uniform1, graphregion(color(white))

for x2
gen normal2 = rnormal(5,2.5)
histogram normal2, normal graphregion(color(white))

I've managed to get separate histos, but to combine i've tried 'graph combine', but do i need to label the graphs to use it in this setting?
also what would be the best way to generate outcome variable for y?

assuming that i can get e using the following
gen normal3 = rnormal(0,1)

is the below the correct way to approach?

generate y=9-1.5*(uniform1)+0.85*(normal2)+normal3

Also, how would i go about renaming variables X1 X2 e and Y to age, screen, sleep and error respectively.

Help would be much appreciated

[Jeph Herrin]

Sounds like a homework problem. Here are some tips:

. help graph combine
. help rename

It's not at all clear why you think

. gen y = 9-1.5*(uniform1)+0.85*(normal2)+normal3

is the correct way to achieve

y = 25 - 2(x1) + 0.95(x2) + e

so you may want to provide more context.

cheers,
J