How to split a matrix?

Hello everyone,
I have the matrix A as follow:
A = [25,240,15,H;25,255,10,R;25,265,20,H;63,240,10,W;63,250,35,L];
The first column in matrix A is the ID that indicates that which row is belongs to what ID (e.g. the first,second and tird row are belongs to ID = 25 , and, forth and fifth row are belong to ID = 63.)
So, I want matrix B like to follow:
B = [ID 240 245 250 255 260 265 270 275 280
25 H H H R R H H H H
63 W W L L L L L L L];
Each DU is divided by 3 (e.g. 15/3 = 5) and then by each of these 5 is added to the ST (e.g. 240, 245 (=240+5), 250 (=245+5).

1 Comment

Moe
Moe on 18 Feb 2016
Characters H, R, W and L can change with any numbers as well. For example, instead of H = 100, R = 101, W = 102, L = 103.

Sign in to comment.

 Accepted Answer

Andrei Bobrov
Andrei Bobrov on 19 Feb 2016
Edited: Andrei Bobrov on 21 Feb 2016
n = xlsread('try1.xlsx',1,'A2:D6');
[a1,~,c1] = unique(n(:,1));
c3 = (n(:,3)-240)/5+1;
c4 = n(:,end)/5;
n2 = n(:,2);
z = accumarray(c1,c4);
b = cumsum(accumarray([c3,c1],1,[z(1),max(c1)]));
out1 = n2(bsxfun(@plus,b,cumsum([0,b(end,1:end-1)])))';
out = [nan,240+(0:z(1)-1)*5;[a1,out1]]

5 Comments

Moe
Moe on 19 Feb 2016
Thanks Andrei, but the code is not working!
Typo :), corrected.
Moe
Moe on 21 Feb 2016
Edited: Moe on 21 Feb 2016
Thanks Andrei,
I tried your code with a larger database, but results aren't correct! I've attached an input file (try2). For example, after running code, look into ID 271, and ST 410. According to excel file the Code should be 103 until ST 1135, but in Matlab output code is 101 still.
Also, I used "ceil" function to get positive integer subscripts for c3 and c4.
Can you please have a look to the code with new data? Thanks in advance!
n = xlsread('try2.xlsx',1);
[a1,~,c1] = unique(n(:,1));
c3 = ceil((n(:,3)-240)/5+1);
c4 = ceil(n(:,end)/5);
n2 = n(:,2);
z = accumarray(c1,c4);
b = cumsum(accumarray([c3,c1],1,[z(1),max(c1)]));
out1 = n2(bsxfun(@plus,b,[0,b(end,1:end-1)]))';
out = [nan,240+(0:z(1)-1)*5;[a1,out1]]
I corrected answer and corrected your data (see try2_1.xlsx)
Moe
Moe on 22 Feb 2016
Grear Andrei, Thanks!

Sign in to comment.

More Answers (0)

Asked:

Moe
on 18 Feb 2016

Commented:

Moe
on 22 Feb 2016

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!