Merging two dissimilar matrix based on common row element

Asked by Poulomi on 4 Aug 2017

Latest activity Answered by Jan on 5 Aug 2017

18 views (last 30 days)

I have two arrays

A = 1  2.576
 2.822
2.679
2.786
2.653
2.596
2.565
B= 1 nan
 nan
 nan
 nan
 nan
 nan
..  nan
..  nan
nan
nan
nan
nan
nan
nan
..  nan
..  nan
nan
nan
nan
..  nan
..  nan
..  nan
nan
nan
nan

I have to create a third matrix C, which have all variables combined based on first column of B, while putting the values of A into it. The empty element, for which no values of A is available should be filled with nan. Desired output is matrix C.

C = 1  2.576
 2.822
 nan
 nan
 nan
 nan
 nan
 nan
 nan
 nan
 nan
 nan
..  nan
..  nan
nan
nan
nan
2.679
2.786
nan
nan
nan
..  nan
..  nan
nan
nan
nan
2.653
2.596
2.565

I tried with intersect, however, it's not working for this case since the common column contains repeated values. Any help?

0 Comments

2 Answers

Link

Answer by Andrei Bobrov on 4 Aug 2017

Edited by Andrei Bobrov on 4 Aug 2017

A = [1  2.576
1   2.822
13  2.679
13  2.786
25  2.653
25  2.596
25  2.565];
B = [kron((1:25)',[1;1;1]),nan(25*3,1)];
C = B;
p = findgroups(A(:,1));
AA = [A(:,1),cell2mat(accumarray(p,1,[],@(x){(1:numel(x))'}))];
q = findgroups(B(:,1));
BB = [B(:,1), cell2mat(accumarray(q,1,[],@(x){(1:numel(x))'}))];
C(ismember(BB,AA,'rows'),2) = A(:,2);

2 Comments

Poulomi Ganguli on 5 Aug 2017

Thanks! it works well.

Jan on 5 Aug 2017

@Poulomi Ganguli: If this solves your problem, please accept the answer.

Link

Answer by Jan on 5 Aug 2017

Or with simple loops:

C  = B;
kC = 1;
nC = size(C, 1);
for iA = 1:size(A, 1)
   a = A(iA, 1);
   for iC = kC:nC
     if C(iC, 1) == a
       C(iC, 2) = A(iA, 2);
       kC       = iC + 1;
       break;   % Stop for iC loop
     end
   end
end

You are now following this question

Poulomi (view profile)