Random number gerator problem

The question that i am working on is as follows

Write a function called randomness that takes three input arguments: limit, n, and m, in that order. The function returns an n-by-m matrix of uniformly distributed random integers between 1 and limit inclusive. You are not allowed to use randi, but you can use rand. You will also find the built-in function floor useful for obtaining integers. To make sure that your result is indeed uniformly distributed, test the output of the function by using the built-in function hist, which plots a histogram.

My code is as follows,

function [r] = randomness(limit,n,m)

r= 1+floor(rand(n,m).*(limit-1))

hist(randomness, 1:limit)

end

It keeps giving me an error for the arguments (20,500000,1) and i am not sure why.I have tried using ceil and fix instead of floor, but to little or no avail.

Answer 1

[James Tursa]

hist(r, 1:limit);

You inadvertently used randomness in your hist call, so MATLAB tried a recursive call with no arguments, hence the error.

You should re-check your algorithm, however, since it currently doesn't produce any values at "limit" and the instructions say "inclusive".

[Sanwal Yousaf]

basically the arguments that i am testing it at are (20,500000,1)because that is where it breaks down. The limit in this case would be 20. WHen i plot this histogram, it comes up something like this which is indicative of the fact that not every number has equal probability of popping up with the same occurence.

[James Tursa]

You should not expect a random sampling to produce the exact number of samples in each bin. But for a large enough sample you can expect the numbers to be reasonably close. There are other tests you can apply to see if your results are in fact "reasonably close", but I think all your instructor wanted was a quick visual check hence the suggestion to use hist. In that sense your chart looks like a pretty good result to me EXCEPT THAT 20 IS MISSING! So, as stated above, your current algorithm is not producing numbers from 1 to 20 "inclusive". You need to look into the reason for this and fix your algorithm. (It's a relatively easy fix ... but I will let you find it).