# split a matrix into two matrices according to some rule

4 views (last 30 days)
Alberto Acri on 12 Feb 2024
Commented: Cris LaPierre on 12 Feb 2024
How can I generate two separate matrices, one containing the coordinates of line A and the other of line B?
figure
plot3(matrix(:,1),matrix(:,2),matrix(:,3),'k.','Markersize',5);
axis equal
grid off
I wanted to try this way but it does not seem the best way.
% range
xRange = [xmin xmax];
yRange = [ymin ymax];
zRange = [zmin zmax];
% indices of points in the range
idx = matrix(:,1) >= xRange(1) & matrix(:,1) <= xRange(2) & ...
matrix(:,2) >= yRange(1) & matrix(:,2) <= yRange(2) & ...
matrix(:,3) >= zRange(1) & matrix(:,3) <= zRange(2);
select_matrix = [matrix(idx,1), matrix(idx,2), matrix(idx,3)];
Alberto Acri on 12 Feb 2024
Hi Stephen! sorry, I realised I had not installed 'Image Processing Toolbox'!
Alberto Acri on 12 Feb 2024
I wanted to know if there is a way to get the subdivision! I don't have any 'rules' at the moment!

Cris LaPierre on 12 Feb 2024
Edited: Cris LaPierre on 12 Feb 2024
I would look into clustering. Here is an attempt that uses dbscan, a spectral clustring algorithm included in the Statistics and Machine Learning Toolbox.
You can learn more about this and other clustering techniques in our Practical Data Science with MATLAB specialization on Coursera. It's free to enroll. Here is a link to the video Introdcution to Clustering Algorithms.
% view the raw data
plot3(matrix(:,1),matrix(:,2),matrix(:,3),'k.','Markersize',5);
% use dbscan to identify clusters
idx = dbscan(matrix,1.5,5);
gscatter(matrix(:,1),matrix(:,2),idx)
Cris LaPierre on 12 Feb 2024
If you know the number of clusters already, use spectralcluster.
idx = spectralcluster(matrix,2);
gscatter(matrix(:,1),matrix(:,2),idx)

### Categories

Find more on Hierarchical Clustering in Help Center and File Exchange

R2021b

### Community Treasure Hunt

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

Start Hunting!