make a cluster from scatter point

Hi

I have attached my scatter poit figure and I want to make a 2 or 3 clusters from them.

This is my code;

for i = 1:size(MeanEchos,1);

for j = 1:size(MeanEchos,2);

scatter(blob_area{2,j},MeanEchos{2,j},'x');

set(gca,'xscale','log');

grid on;

xlabel('Blob Area');

ylabel('Mean Echo Intensity');

xlim([0 10^5]);

ylim([0 5])

end

end

I would be greateful if anyone would be able to help me.

Thanks in advance

Answer 1

[Ameer Hamza]

If you have statistical and machine learning toolbox, then try kmeans(): https://www.mathworks.com/help/releases/R2020b/stats/kmeans.html

[Fatemeh Shomal Zadeh]

Thanks! how I can make 1 array from these two double cell array blob_area{2,j} and MeanEchos{2,j}?

because when I want to use kmeans, it needs to be one array.

blob area here is x of my figure and means echos is y.

[Ameer Hamza]

I think following should work

x1 = [blob_area{2,:}].';
x2 = [MeanEchos{2,:}].';
X = [x1 x2];

Input X to kmeans(). If there is an error, can you attach the variables blob_area and MeanEchos in a .mat file.

[Fatemeh Shomal Zadeh]

Thanks so much for your respond. I do really appreciate it.

With the new code that you ahve sent, I am receiving an error of 'Vectors must be the same length.'

I have attached my data here. Baically as you can see in the screen shot od scatter point, the first distribution needs to be cluster #1 and second distribution needs to be cluster #2.

[Ameer Hamza]

These lines wok without error on my machine. How are you typing these lines?

[Fatemeh Shomal Zadeh]

Yes you are right. I used your code and thanks so much for that.

However, I want to make sure that I am receieving the right resualt.

My all lines are stated as bellow;

x1 = [blob_area{3,:}].';

x2 = [MeanEchos{3,:}].';

X = [x1 x2];

opts = statset('Display','final');

[idx,C] = kmeans(X,2,'Distance','cityblock',...

'Replicates',255,'MaxIter',1000,'Options',opts);

figure;

plot(X(idx==1,1),X(idx==1,2),'r.','MarkerSize',12)

hold on

plot(X(idx==2,1),X(idx==2,2),'b.','MarkerSize',12)

plot(C(:,1),C(:,2),'kx','MarkerSize',15,'LineWidth',3)

legend('Cluster 1','Cluster 2','Centroids',...

'Location','NW')

title 'Cluster Assignments and Centroids'

hold off

I attached the figure as well. based on the initial figure, is my result right?

Also, thanks so much for all your help, with that 2 lines I got my answer after 2 days working.

[Ameer Hamza]

From the images, it seems that the results are correct. It clearly shows two clusters. I am glad to be of help!

[Fatemeh Shomal Zadeh]

Great. Thanks a lot!