Create a random array with certain requirements

Hi,

I would like to create an array V that contains random elements (numbers) with the following characteristic:

1) The array should have lenght equal to X

2) The avarage value of the elements inside the array should be Y

3) The elements should stay inside an interval T= [A to B]

4) The elements should not necessary be integer, could also be decimal till up to the fourth decimal place.

Let's consider the following example:

GIven as inputs: X= 10 , Y=19.9948 , T= [0 to 50]

I would like to obtain a random array like

V = [1 5.5674 28.9764 18 30.8433 35.3389 4.6333 10 16 49.5887]

Thanks for your help

[luca]

With something like that I Create a vector of 10 random values drawn from a normal distribution with a mean of 15 and a standard deviation of 20.But some values are out of the range [0 50]. As it possible to see from y.

a = 20;

b = 15;

y = a.*randn(10,1) + b;

Do you know how I can change the code, so that it works?

Answer 1

[Muthu]

Hello Luca,

I have tried to draft a program which does just the requested function.

You can still expand and fine tune this program with more condtions.

function res_array = random_array(x, y, interval)
temp_array = rand(1,x) * interval(2);
temp_mean = mean(temp_array);
add_to_sum = x * (y-temp_mean);
add_in_each_term = add_to_sum / x ; 
res_array = temp_array + add_in_each_term ;
end

result = random_array(10,19.9948,[0,50])

result =

   37.1853    7.8636   35.8637    7.3257   41.6126   12.6486    4.9792    7.7036   25.9517   18.8139

[luca]

May you explain how could I use the function? I cannot found the result on my PC

[luca]

And what is "n"?

[luca]

I suppose n is equal to X. In any case, with this code sometimes I obtain something negative like this:

  Columns 1 through 8

   30.4070    2.6201   27.5111    2.3479   11.5966   24.4533   32.1837   -2.7714

  Columns 9 through 10

   39.6417   31.9580
   
   

Do you know how to solve this problem?

[Muthu]

n is X, its a mistake.

The negative value is because of the subtraction from Y:

add_to_sum = x * (y-temp_mean);
add_in_each_term = add_to_sum / x ; 
res_array = temp_array + add_in_each_term ;

when temp_mean is more than Y, we get negative value to add to sum.

hence that in turn is subtracted from all terms in res_array. When the random number generated is less than the value subtracted, we get negative i suppose.

It is a very good point to start adding conditions further in the function to get the proper result array.