New to MatLab: How to enter given data for a random variable (&find /fit distribution function)

Hello,

I am very new to MatLab so I am still struggling a lot.

What I have given:

I have a random variable X with thoutcomes and probabilites.

How do I put this data in MatLab? I know this is very basic but I am struggling with it. In the next step I need to find and draw a fitting distribution function, I think I know how to do this as I found some tutorials, but all these tutorials only used sample data sets and not data sets they had to put in theirselves.

Thanks already!

Answer 1

[the cyclist]

If you have the Statistics and Machine Learning Toolbox, you can use the randsample function:

N = 100;
outcome = [1; 3; 4; 5; 9; 10];
prob = [0.20; 0.15; 0.10; 0.05; 0.35; 0.15];
y = randsample(outcome,1000000,true,prob);

will give 100 samples from those outcomes, weighted by the probabilities.

[Lisa Maria Teppich]

Thanks for your answer! I dont have this/ am not sure whether I am allowed to use if for the assignment. Is there another way to do so without it?

Answer 2

[Torsten]

n = 30;
sample = random(n)

function sample = random(n)
  uniform = rand(n,1);
  sample = zeros(size(uniform));
  for i = 1:n
    v = uniform(i);
    if v <= 0.2
        s = 1;
    elseif v > 0.2 && v <= 0.35
        s = 3;
    elseif v > 0.35 && v <= 0.45
        s = 4;
    elseif v > 0.45 && v <= 0.5
        s = 5;
    elseif v > 0.5 && v <= 0.85
        s = 9;
    else
        s = 10;
    end
    sample(i) = s;
  end
end

Answer 3

[the cyclist]

Here is a method that does not require the Statistics and Machine Learning Toolbox:

% Input data
N = 100;
outcome = [1; 3; 4; 5; 9; 10];
prob = [0.20; 0.15; 0.10; 0.05; 0.35; 0.15];

% Algorithm
r = rand(1,N);
index = numel(outcome) - sum(r <= cumsum(prob)) + 1;
X = outcome(index);