How to create a random Bernoulli matrix ?

I want to create a 256x256 random Bernoulli matrix, how to do that in matlab ?

[Bilal Siddiqui]

It's simple. A Bernoulli trial produces one of only two outcomes (say 0 or 1). You can use binord. For example p=0.2; n=256; A=binornd(1,p*ones(n));

produces an nxn array of Bernoulli trials which are either 0 or 1 in each outcome. Hope this answers your question.

Answer 1

[Robert Dylans]

p=0.5;        %probability of success
n=256;
A=rand(n);
A=(A<p)

or

p=0.5;
A=(rand(256)<p)

[Mohab Mostafa]

Thank you so much, another question (if it's fine with you), how to generate a random Gaussian matrix ? (also 256x256)

[Robert Dylans]

A=normrnd(mean,standard dev,m,n);

For a cluster around a an average of '0' with sigma 1:

A=normrnd(0,1,256,256);

[Mohab Mostafa]

and what if I want a -+1 Bernoulli matrix to be used in compressed sensing as a sensing matrix ?

[Walter Roberson]

Careful, probability 1/2 might not be! See http://uk.mathworks.com/matlabcentral/answers/246999-probability-1-2-with-matlab-random-number-generators

[Walter Roberson]

If you have a matrix A of 0 and 1 then you can make it a matrix of -1 and +1 by using

A*2 - 1

Answer 2

[Byron McMullen]

I use this.

bernoulli_matrix = randsrc(rows,cols,[-1,1]);

[Weijian Liu]

In my opinion, this is the best one.

Answer 3

[Bilal Siddiqui]

It's simple. A Bernoulli trial produces one of only two outcomes (say 0 or 1). You can use binord. For example p=0.2; n=256; A=binornd(1,p*ones(n));

produces an 256x256 array of Bernoulli trials which are either 0 or 1 in each outcome. Hope this answers your question.