skewness in random numbers

I am generating 30 random numbers ranging within 8 32 by

w1=randi([8 32],1,30);

Now I need a skewness in this randomly generated..i.e. the avg difference between the elements needs to be smaller,

like 8 8 9 10 10 11 12 13 13 14......like this way

If I could draw this distribution from a normal distribution then the std deviation that's I need to make smaller...

is there any way to do so?

Answer 1

[Samayochita]

Hi Tusu,

I understand that you want to generate random numbers between 8 and 32, but instead of the numbers being completely random, they should be closer together (less variation). You can control the spread of randomly generated numbers by drawing from a normal distribution using ‘randn’

(https://www.mathworks.com/help/matlab/ref/randn.html).

You can follow the steps below to achieve this:

1. Generate numbers with a small standard deviation to keep them close to the mean.
2. Adjust the mean and standard deviation to fit within the desired range [8, 32].
3. Ensure the values stay in the range [8, 32] by rounding and clamping.

I have provided the sample code below, additionally you can sort the values before displaying them using ‘sort’

(https://www.mathworks.com/help/matlab/ref/double.sort.html) to ensure they are in ascending order (as shown in your example).

n = 30;        % Number of random values
mu = 20;       % Mean of the distribution (center of the range)
sigma = 3;     % Small standard deviation for less spread
w1 = round(mu + sigma * randn(1, n)); % Generate numbers using normal distribution

% Clamp values within [8, 32]
w1(w1 < 8) = 8;
w1(w1 > 32) = 32;

% Optional -> sort in ascending order
w1 = sort(w1);

% Display the result
disp(w1);

I hope this works for you.