Create a single variable to store multiple correlation coefficients

I have a 300x7 matrix (A) and a 1x300 row vector (B). How do I calculate the CC for each of the 7 columns in A against the single row in B and store them in a single variable?

Answer 1

[John D'Errico]

A = rand(300,7);
B = rand(1,300);

C = corrcoef([A,B']);
C = C(8,1:7)

C = 1×7

0.0055 -0.0433 0.0186 -0.0125 0.0854 0.0879 -0.1811

The correlation coefficiens of each column of A, against B are given in C; Admittedly, if the array A had many columns, then the above is wasteful in terms of CPU cycles. So I could also have done this:

C2 = (B-mean(B))/norm(B-mean(B))*normalize(A - mean(A),1,'norm')

C2 = 1×7

0.0055 -0.0433 0.0186 -0.0125 0.0854 0.0879 -0.1811

As you can see, the results are the same.

Answer 2

[Luca Ferro]

this would do the trick and store the correlation values in a 300x7 matrix

for rr=1:size(AA,1)
    for cc=1:size(AA,2)
        CC(rr,cc)=corrcoef(A(rr,cc),B(1,rr));
    end
end

The order of the matrix follows the order of A, meaning that CC(1,1) is the correlation of A(1,1) and B(1), and so on