MATLAB Merging Arrays - arrays

Question

I am unable to figure out how to merge two arrays. My data is like this with arrays A and B.
A = [ 0 0; 0 0; 2 2; 2 2;]
B = [ 1 1; 1 1; 3 3; 3 3; 4 4; 4 4; 5 5; 5 5;]
and I need the final array "C" to look like this after merging:
C = [ 0 0; 0 0; 1 1; 1 1; 2 2; 2 2; 3 3; 3 3; 4 4; 4 4; 5 5; 5 5;]
I've tried using different ways with reshaping each array and trying to use a double loop but haven't got it to work yet.
In my actual data it is inserting 9 rows of array B following 3 rows of array A and then repeated 100 times. So, there are 12 new merged rows (3 rows from array A and 9 rows from array B) repeated 100 times with a final row number == 1200. Array A actual data has 300 rows and actual Array B data has 900 rows
thanks,

Answer 1

Here's a solution using only reshape:
A = [ 6 6; 3 3; 5 5; 4 4;]
B = [ 0 0; 21 21; 17 17; 33 33; 29 29; 82 82;]
A_count = 2;
B_count = 3;
w = size(A,2); %// width = number of columns
Ar = reshape(A,A_count,w,[]);
Br = reshape(B,B_count,w,[]);
Cr = [Ar;Br];
C = reshape(Cr,[],w)
The [] in reshape means "how ever many you need to get to the total number of elements". So if we have 12 elements in B and do:
Br = reshape(B,3,2,[]);
We're reshaping B into a 3x2xP 3-dimensional matrix. Since the total number of elements is 12, P = 2 because 12 = 3x2x2.
Output:
A =
6 6
3 3
5 5
4 4
B =
0 0
21 21
17 17
33 33
29 29
82 82
C =
6 6
3 3
0 0
21 21
17 17
5 5
4 4
33 33
29 29
82 82

Answer 2

Approach #1
This could be one approach assuming I got the requirements of the problem right -
%// Inputs
A = [ 6 6; 3 3; 5 5; 4 4;];
B = [ 0 0; 21 21; 17 17; 33 33; 29 29; 82 82;];
%// Parameters that decide at what intervals to "cut" A and B along the rows
A_cutlen = 2; %// Edit this to 3 for the actual data
B_cutlen = 3; %// Edit this to 9 for the actual data
%// Cut A and B along the rows at specified intervals into 3D arrays
A3d = permute(reshape(A,A_cutlen,size(A,1)/A_cutlen,[]),[1 3 2])
B3d = permute(reshape(B,B_cutlen,size(B,1)/B_cutlen,[]),[1 3 2])
%// Vertically concatenate those 3D arrays to get a 3D array
%// version of expected output, C
C3d = [A3d;B3d]
%// Convert the 3D array to a 2D array which is the final output
C_out = reshape(permute(C3d,[1 3 2]),size(C3d,1)*size(C3d,3),[])
Sample run -
A =
6 6
3 3
5 5
4 4
B =
0 0
21 21
17 17
33 33
29 29
82 82
A_cutlen =
2
B_cutlen =
3
C_out =
6 6
3 3
0 0
21 21
17 17
5 5
4 4
33 33
29 29
82 82
Approach #2
Just for the love of bsxfun, here's one approach with it and ones (no reshape or permute) -
%// Assuming A_cutlen and B_cutlen decide cutting intervals for A and B
%// Concatenate A and B along rows
AB = [A;B]
%// Find the row indices corresponding to rows from A and B to be placed
%// according to the problem requirements
idx1 = [1:A_cutlen size(A,1)+[1:B_cutlen]]
idx2 = [A_cutlen*ones(1,A_cutlen) B_cutlen*ones(1,B_cutlen)]
idx = bsxfun(#plus,idx1(:),idx2(:)*[0:size(A,1)/A_cutlen-1])
%// Re-arrange AB based on "idx" for the desired output
C = AB(idx,:)

Answer 3

based on your new criteria this is what you want. My solution isn't the nicest looking (maye someone can think of a nice vectorized approach), but it works
a_step = 2;
b_step = 3;
C = zeros(size([A;B]));
%we use two iterators, one for each matrix, they must be initialized to 1
a_idx = 1;
b_idx = 1;
%this goes through the entire c array doing a_step+b_step rows at a
%time
for c_idx=1:a_step+b_step :size(C,1)-1
%this takes the specified number of rows from a
a_part = A(a_idx:a_idx+a_step-1,:);
%tkaes the specified number of rows from b
b_part = B(b_idx:b_idx+b_step-1,:);
%combines the parts together in the appropriate spot in c
C(c_idx:c_idx + a_step + b_step -1,:) = [a_part;b_part];
%advances the "iterator" on the a and b matricies
a_idx = a_idx + a_step;
b_idx = b_idx + b_step;
end
using
A = [ 6 6; 3 3; 5 5; 4 4;]
B = [ 0 0; 21 21; 17 17; 33 33; 29 29; 82 82;]
produces
C =[6 6; 3 3; 0 0; 21 21; 17 17; 5 5; 4 4; 33 33; 29 29; 82 82;]