combine two double cells into 1

i have two Double[] cells with size of 1*n; every element is has some numbers in it . like :

C1{1} = [1 2 3 4]
C2{1} = [5 2 3 7]
.
.
.
C1{n} = [1 2]
C2{n} = [4 5]

how can i have a cell like this :

C{1} = [1 2 3 4 5 7]
.
.
.
C{n} = [1 2 4 5]

i`m totally new to matlab and sorry if its already been answered , but i search a lot and couldn`t find the answer .

[Stephen23]

"two Double[] cells"

A cell array is NOT a double array (but can contain double arrays).

[Hamid Salari]

@Stephen thanks . i didn`t know that .

Answer 1

[Stephen23]

>> C1{1} = [1,2,3,4];
>> C2{1} = [5,2,3,7];
>> C1{2} = [1,2];
>> C2{2} = [4,5];
>> Z = cellfun(@(a,b)unique([a,b]),C1,C2,'uni',0);
>> Z{:}
ans =
     1     2     3     4     5     7
ans =
     1     2     4     5

Answer 2

[Jan]

If you are in doubt, start with a loop:

C1 = {[1 2 3 4], [1 2]};
C2 = {[5 2 3 7], [4 5]};

C = cell(size(C1));
for iC = 1:numel(C)
  C{iC} = unique([C1{iC}, C2{iC}]);
end

[Stephen23]

+1 Note that the example has only the unique values (so is not just the concatenation):

C{iC} = unique([C1{iC}, C2{iC}]);

[Jan]

Thanks, Stephen. I hit the Submit button too early.

[Hamid Salari]

Thanks Jan.

Answer 3

[Star Strider]

This works, and should work with your entire ‘C1’ and ‘C2’:

C1{1} = [1 2 3 4];
C1{2} = [1 2];
C2{1} = [5 2 3 7];
C2{2} = [4 5];

C3 = cellfun(@union, C1, C2, 'Uni',0);
C3{1}                                       % Display Result
C3{2}                                       % Display Result

ans =
     1     2     3     4     5     7

ans =
     1     2     4     5

EDIT — Added output (the ‘ans’ variables).

[dpb]

Good use for UNION, SS... +1

Answer 4

[dpb]

That's a little towards the advanced-beginner side... :)

>> C=cellfun(@(c1,c2) unique([c1 c2]),C1,C2,'uniform',0);
>> C{:}
C =
   1     2     3     4     5     7
C =
   1     2     4     5
>>

The @ symbol is defining an "anonymous function", with two arguments; cellfun passes the content of its two cell array arguments to the defined function cell-by-cell and unique does what it sounds like which is to return the unique values in its argument; the [] simply concatenate the two vectors.

[Hamid Salari]

Thanks for "anonymous function" . i will look it up . seems handy .

[Jan]

Anonymous functions are very handy, but tend to be slow e.g. in cellfun.

[dpb]

Note Star S's use of UNION over UNIQUE, though...saves a step of the concatenation.

I've never done specific timing; where the syntax is easy I'll write the CELLFUN solution as above and only if it turns out to be a bottleneck worry about what generally would be small differences in run time. Then again, I'm not actively consulting any longer so don't in general ever have large datasets where it would ever be an issue so "caveat emptor" reigneth, methinks. Keep Jan's note in mind if you're still waiting next week for the prompt to come back... :)