How do I generate a random number with normal distribution

I want to generate a random matrix with normal distribution between -10 and 10. I cannot use randn() because it has no boundaries. Is there any other to generate such matrix??

Answer 1

[Star Strider]

This is called a truncated distribution. While you are correct that it is no longer strictly a normal distribution (that extends to infinity to either side of the mean), you can nevertheless define it. The extended discussion is in the truncate (link) documentation section in Truncate a Probability Distribution (link).

For your Question, this would be:

pd = makedist('Normal')
t = truncate(pd,-10,10)
r = random(t,10000,1);

Experiment to get the result you want.

[Jamie Derulo]

Thanks. It really worked. However I found out that the mean and standard deviation has effects to make me get close to the boundary -10 and 10. So i made a simple loop to check which best fits.

for i = 1:9
    pd = makedist('Normal',0,i);
    t = truncate(pd,-10,10);
    r = random(t,10000,1);
    subplot(3,3,i)
    hist(r,100)
    max(r)
    min(r)
end

In here, I set my mean to 0 and my standard deviation is i. When i is >= 3, it fits perfectly for what I am looking for as my max and min are more close to my boundaries.

Thanks again for the help

[Star Strider]

As always, my pleasure!