Combining two different matrices of different sizes

I want to combine two different matrices with a different number of columns so that the output has the lower number of columns. I also want the columns that were "modified" to have the average value of what they were modified. I think these examples clarify what I'm trying to say:

For example, for matrices that are multiples of each other:

 A=[5 6]
 B=[1 2 3 4]
I would combine these matrices so
 C=[5 6; 1.5 3.5]

Now if the matrices aren't multiples of each other, then it would combine to whichever value it finds fit, for example:

 A=[5 6]
 B=[1 2 3 4 5]
Then the output is:
 C=[5 6; 2 4.5] (The first 3 values were averaged)

Thanks

[Azzi Abdelmalek]

What is the size of A?

[Jan]

The question is not clear. What happens for other number of columns for A and C?

Answer 1

[Azzi Abdelmalek]

A=[5 6] 
B=[1 2 3 4 5]
m=numel(B);
m1=ceil(m/2)
C=[A;mean(B(1:m1)) mean(B(m1+1:end))]

[Amith Kamath]

doesn't this work only with numel(A) == 2 ?

[Azzi Abdelmalek]

Try this

A=[5 6 7] 
B=[1 2 3 4 5 7 8 6 9 10 11]
n=numel(A)
m=numel(B);
m1=floor(m/n)
m2=m-m1*n+m1
C1=mean(B(1:m2))
for k=m2+1:m1:m
C1=[C1;mean(B(k:k+m1-1))]
end
C=[A;C1']

%or I prefer

A=[5 6 7 80 47] 
B=[1 2 3 4 5 6 7 8  9 10 11 12 13 14 15  25  35]
n=numel(A);
m=numel(B);
m1=floor(m/n);
h=zeros(1,n);
h(1:m-m1*n)=1
v2=cumsum(ones(1,n)*m1+h);
v1=[1 v2(1:end-1)+1];
C=[A;arrayfun(@(x) mean(B(v1(x):v2(x))),1:n)]

%In the first code the mean values were calculated for

[B(1:5) B(6:8) B(9:11)]

% In the second code

[B(1:4) B(5:8) B(9:11)]