# random number generation

Brian Blais bblais at bryant.edu
Tue Aug 17 02:57:29 CEST 2010

```On Aug 16, 2010, at 20:37 , Jah_Alarm wrote:

> hi,
>
> I need to generate a binary array with a specified average proportion
> of 1s (e.g. [1 0 0 0
>
> 0 1 0 0] has this proportion = 25%). In Matlab I run something like
> random(m,n)<p where p is the value
>
> between 0 and 1. I'm trying to use random.randint(0,2,size=[m,n]), but
> I don't understand how to specify this proportion p.
>

if you're coming from matlab, then you should use the numpy package
(and you can post questions on the numpy list).  In that case, you
can do:

from numpy import *
random.rand(5,7)<0.25

array([[False,  True,  True, False, False,  True, False],
[False, False, False, False,  True, False, False],
[ True, False, False, False, False, False,  True],
[ True, False,  True, False, False, False, False],
[False, False, False, False, False, False, False]], dtype=bool)

just like matlab.

bb

--
Brian Blais
bblais at bryant.edu
http://web.bryant.edu/~bblais
http://bblais.blogspot.com/

```