binary random variable generator

[Vivian Phan]

Hi all

I want to generate binary random sample with the mean 0.3.
I think that runiformint(0,1) might help me but the mean is 0.5.

I don't know how to do it. Please help!
Thank you

[Nick Cox]

gen wanted = runiform() < 0.3

[Vivian Phan]

Thank you Nick. It helps me.

[Maarten Buis]

Also see: M. Buis (2007) Stata Tip 48: Discrete uses for uniform(). The Stata Journal, 7(3): 434-435.
http://www.stata-journal.com/article...article=pr0032

Note that at the time that that Stata tip was written runiform() was called uniform(). So whenever you see uniform() in that article, you need to write runiform()

[Vivian Phan]

Thank you, Maarten. It is really helpful.

[Felix Stips]

The above code does not assure that the mean of wanted will be 0.3, which of course will only be possible for specific numbers of observations in the dataset.

sysuse auto.dta
gen wanted = runiform() < 0.3
sum wanted

However, I think this code gives a better approximation than the first version

set seed 2803
gen random = runiform()
sort random
local cut = round((_N / 10) * 3)
gen wanted2 = _n <= `cut'
sum wanted2

[Maarten Buis]

It depends on what you want. If you want to get a mean that every time you run it is as close to .3 as possible, then what Felix Stips proposes is an idea. However, if you want to get a set of draws from a bernouilie distribution, then what I propose is the right thing.