Check an entire matrix against each value of another matrix

Dear all,

I have the following code;

1. I want to check matrix b against each value of matrix a, and extract values in matrix b based on a condition.
2. The condition when delta 1 == 0, the value of matrix b corresponding to 0 is extracted into c.

However, I am getting an error Index in position 1 exceeds array bounds (must not exceed 1) in the line below. I have 2 questions:

1. Why is this error happening?
2. Is there another way to accomplish this without using loops?

Thank you.

a = [4 3];
b = [1 4.2 5 6 7 3.2];
c = zeros(size(a));

for i = 1:length(a)
    delta1(i,:) = abs((b-a(i))/a(i));
    for j = 1:length(delta1)
        if delta1(i,j) < 0.3
           c(i) = b(i,j);               % This line
        end
    end
end

Answer 1

[Jon]

If I understand what you are looking for, I think this will do it in a very simple way

c = intersect(a,b)
>> a = [4 3];
b = [1 4 5 6 7 3];
c = intersect(a,b)

c =

     3     4

[EYKL]

Hi @Jon, this is a nice answer. But what if I have different decimal points in both matrix a and b? Apologies, please see the updated question.

[Jon]

Sorry, I based my answer on what I think I saw in your original post, where the values were all integers. Did you edit that since it was originally posted? Anyhow in that case it is a little harder. I will see what I come up with.

[Jon]

a =

    4.0000    0.2000    3.0000

>> b

b =

    1.0000    4.2000    5.0000    6.0000    7.0000    3.2000

>> delta = abs((b -a')./a')

delta =

    0.7500    0.0500    0.2500    0.5000    0.7500    0.2000
    4.0000   20.0000   24.0000   29.0000   34.0000   15.0000
    0.6667    0.4000    0.6667    1.0000    1.3333    0.0667

>> c = a(any(delta < 0.3,2))

c =

     4     3

[Jon]

Ooops, that last line should be

c = b(any(delta < 0.3))
c =

    4.2000    5.0000    3.2000

if you want to list the values that match out of b instead of the ones out of a (as in code above)

[Jon]

You should also check out the MATLAB function ismembertol https://www.mathworks.com/help/matlab/ref/ismembertol.html which is pretty much purpose built for the operation you are trying to perform, but the error tolerance is a little different. If you are comfortable with their way of computing the tolerance you can use this directly