# how do i generate a 1X4 matrix with until the sum of the matrix =100

27 views (last 30 days)
Austin on 5 Nov 2013
Answered: Image Analyst on 5 Nov 2013
how do i generate a 1X4 random matrix until the sum of the matrix equals to 100
with the following condition: (each time the matrix is generated, the last element of the newly generated 1X4 matrix being the first element of last generated 1X4 matrix?) (advise using persistent variables)

Azzi Abdelmalek on 5 Nov 2013
What kind of random numbers, integers? real? its maximum value?
Austin on 5 Nov 2013
sorry, its real. no maximum value given, each time the sum of the 4 elements does not add up to 100,it regenerates 4 real numbers again..
a=100*rand(1,4) if sum(a)~=100
.....

Azzi Abdelmalek on 5 Nov 2013
The probability to fit your conditions is very small, you can run your code all the year, it's not obvious to get the result.

Image Analyst on 5 Nov 2013
Try this:
clc;
maxLoops = 1000
loopCounter = 1;
while loopCounter <= maxLoops
theArray = randi(97, 1, 4);
value = sum(theArray);
if value == 100
fprintf('The array = [%d, %d, %d, %d], found after %d tries.\n',...
theArray(1), theArray(2), theArray(3), theArray(4), loopCounter);
break;
end
loopCounter = loopCounter + 1;
end
if loopCounter >= maxLoops % The failsafe kicked in
end

Image Analyst on 5 Nov 2013
By the way, persistent variables are not needed at all.
Azzi Abdelmalek on 5 Nov 2013
In his comment, he asks for real numbers
Image Analyst on 5 Nov 2013
According to the FAQ http://matlab.wikia.com/wiki/FAQ#Why_is_0.3_-_0.2_-_0.1_.28or_similar.29_not_equal_to_zero.3F we know that cannot happen with real numbers, so I used integers.

Simon on 5 Nov 2013
Hi!
Another way may be to create 4 random numbers, take what is missing upto 100 and add to the numbers:
% random numbers
a = 100*rand(1,4);
% missing part
miss = 100 - sum(a);
% choose one possibility
% add to numbers, possibility 1 (linear)
a = a + miss/4;
% add to numbers, possibility 2 (proportional)
a = a + miss*(a/sum(a))
% add to numbers, possibility 3 (???)

Image Analyst on 5 Nov 2013
Here's a way where you can specify how close you want to get to 100 using floating point numbers. Like Azzi and I said, there's virtually no chance you'll ever hit 100 exactly out to the millionth decimal place. My code below will find numbers that get you to within 0.01, which you can change to get closer if you want.
clc;
maxLoops = 100000
loopCounter = 1;
while loopCounter <= maxLoops
theArray = 97*rand(1, 4);
theSum = sum(theArray);
if abs(theSum - 100) < 0.01
fprintf('The sum = %.5f.\nThe array = [%g, %g, %g, %g], found after %d tries.\n',...
theSum, theArray(1), theArray(2), theArray(3), theArray(4), loopCounter);
break;
end
loopCounter = loopCounter + 1;
end
if loopCounter >= maxLoops % The failsafe kicked in
end