Select matrices with nonzero rows from a bigger matrix ?
3 visualizzazioni (ultimi 30 giorni)
Mostra commenti meno recenti
Geraldine Ejimadu
il 26 Apr 2020
Commentato: dpb
il 27 Apr 2020
Hello guys!
I am struggling with this problem:
I have a big matrix that is made of few zeros raws and many nonzero raws.
This is a fake simple example to explain, because in reality I am working with big data :(:
B= [0 0 0 0;
0 0 0 0;
0 0 0 0;
0 0 0 0 ;
1 1 2 3;
0 0 0 3;
1 1 1 0;
4 4 4 4;
0 0 0 0 ;
0 0 0 0;
0 0 0 0 ;
0 0 0 8;
0 0 1 0;
0 1 1 0;
0 0 0 0;
0 0 0 0 ];
The zero rows in this matrix, indicate that the signal stopped and I just need to analyse the non zero raws in chunks.
Therefore would like to obtain:
chunk_1=[1 1 2 3;0 0 0 3;1 1 1 0;4 4 4 4]
chunk_2=[0 0 0 8;0 0 1 0;0 1 1 0;]
As you can see the chunks don't have to have the same amount of raws, since I analyse them individually, and the zero chunks also don't have the same number of rows.
I hope you guys have more knowledge than ve and could help me!
Thanks in advance !
0 Commenti
Risposta accettata
Ameer Hamza
il 26 Apr 2020
Modificato: Ameer Hamza
il 26 Apr 2020
This nice little function from the image processing toolbox (bwconncomp) will be very useful here.
B = [0 0 0 0;0 0 0 0;0 0 0 0;0 0 0 0 ;1 1 2 3;0 0 0 3;1 1 1 0;4 4 4 4; 0 0 0 0 ;0 0 0 0;0 0 0 0 ;0 0 0 8;0 0 1 0;0 1 1 0;0 0 0 0;0 0 0 0 ];
mask = sum(B, 2)~=0;
comps = bwconncomp(mask);
compMaskList = comps.PixelIdxList;
B_components = cell(numel(compMaskList), 1);
for i=1:numel(compMaskList)
B_components{i} = B(compMaskList{i}, :);
end
This shorter version is equivalent to the above code
comps = bwconncomp(sum(B, 2)~=0);
B_comps = cellfun(@(x) {B(x, :)}, comps.PixelIdxList);
5 Commenti
Ameer Hamza
il 26 Apr 2020
If you run 'sum(B, 2)~=0', you get
>> sum(B, 2)~=0
ans =
16×1 logical array
0
0
0
0
1
1
1
1
0
0
0
1
1
1
0
0
So bwconncomp detects that there are two groups of 1s and gives back their row number.
Più risposte (1)
dpb
il 26 Apr 2020
Any number of these kinds of Q? on Answers -- under "runs" or any number of other terms...but, it's pretty simple to code from scratch--
d=diff([false;any(B,2)]); % locations with any observations--T
ix=[find(d==1) find(d==-1)]; % find start, end of sections 0->1, 1->0
C=arrayfun(@(i) B(ix(i,1):ix(i,2),:),1:size(ix,1),'UniformOutput',false); % cell array of contents by group
4 Commenti
dpb
il 27 Apr 2020
Very common type problem and typical use of diff for pattern-matching. "Trick" worth knowing; can adapt to many situations with thought.
May not always have Image Processing TB (as I don't) so need to be able to do oneself...
Vedere anche
Categorie
Scopri di più su Logical in Help Center e File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!