Normal random number generation

I want to generate some random numbers which should lie between 0 and 1. Also those values should follow nornal ditribution. Is there any readymade function for it, in MATLAB?

[Dyuman Joshi]

You can normalize the data in between 0 to 1, though I am not sure if it will still follow normal distribution or not.

y=randn(1,7)

y = 1×7

-2.3212 -0.6273 -0.4058 0.3280 0.0769 1.5305 -0.0685

y=(y-min(y))/(max(y)-min(y))

y = 1×7

0 0.4398 0.4973 0.6878 0.6226 1.0000 0.5849

I also recommend you check out these comments -

Restricting normal values to a particular range

A Normal distribution doesn't have limits

Answer 1

[Torsten]

pd = makedist('Normal');
t = truncate(pd,0,1);
r = random(t,1e6,1);
figure(1)
histogram(r,100)

Or directly:

r = -sqrt(2)*erfinv(rand(1e6,1)*erf(-1/sqrt(2)));
figure(2)
histogram(r,100)

Of course, the random numbers are not normally distributed in the usual sense.

[Dyuman Joshi]

Quite a neat function.

However, OP might not have access to Stats and ML toolboox.

Answer 2

[Bruno Luong]

Here is an alternatve without stats tbx

https://uk.mathworks.com/matlabcentral/fileexchange/23832-truncated-gaussian?s_tid=srchtitle

X=TruncatedGaussian(-1,[0 1],[1, 1e6]);
histogram(X)