Obsidian | Level 7

## How to simulate a Bernouilli variable

Hi,

I am wondering about simulate a bernouilli variable with parameter h.

I know how to simulate a normale variable for example :

retain seed 7453293 ;

call rannor(seed,epsi);

X_t= epsi;

But I don't know how to simulate a bernouilli.

1 ACCEPTED SOLUTION

Accepted Solutions
SAS Super FREQ

## Re: How to simulate a Bernouilli variable

The RAND function (DATA step) and the RANDGEN function (SAS/IML) supports about 25 distributions, including the Bernoulli distribution.

The code you included looks like you want the DATA step solution, even though you posted to the SAS/IML forum (??): The following DATA step generates 10 random Bernoulli variates, each which has a 0.3 probability of appearing:

``````data Bern;
call streaminit(7453293);
do i = 1 to 15;
x = rand("Bernoulli", 0.3);
output;
end;
run;``````

By the way, there are many reasons why you should transition to use RAND instead of the older RANUNI and RANNOR functions, which are deprecated.

SAS Super FREQ

## Re: How to simulate a Bernouilli variable

The RAND function (DATA step) and the RANDGEN function (SAS/IML) supports about 25 distributions, including the Bernoulli distribution.

The code you included looks like you want the DATA step solution, even though you posted to the SAS/IML forum (??): The following DATA step generates 10 random Bernoulli variates, each which has a 0.3 probability of appearing:

``````data Bern;
call streaminit(7453293);
do i = 1 to 15;
x = rand("Bernoulli", 0.3);
output;
end;
run;``````

By the way, there are many reasons why you should transition to use RAND instead of the older RANUNI and RANNOR functions, which are deprecated.

From The DO Loop