K-means clustering method

I have a set of lightning data I am trying to cluster by k-means. I am using a k-pick plot graph of point to centroid distances to determine the appropriate number of clusters to choose and then setting a for loop to iterate up to 10 times (see below code). The issue that I am running into is that it is plotting less clusters than I am specifying... why is this? Any help would be appreciated.

all_LTG = [LTGlon,LTGlat];%combine the lat/lon LTG data into one matrix
j=4; %k clusters selected from kpickplot
[LTGidx2,LTGC2,LTGsum2,LTGD2] = kmeans(all_LTG,j,'replicates',10,'display','final');
ptsymb = {'bs','r^','md','go','c+','y*','k.'}; %assign symbols and colors for different clusters
for i=1:j
    LTGclust = find(LTGidx2==i); 
    plot(all_LTG(LTGclust,1),all_LTG(LTGclust,2),ptsymb{i}); 
    hold on 
end
plot(LTGC2(:,1),LTGC2(:,2),'ko'); 
plot(LTGC2(:,1),LTGC2(:,2),'kx'); 
hold off
title('LTG: Cluster Plot') 
xlabel('Longitude º ')
ylabel('Latitude º ') 
xlim([-35 55]);
ylim([-40 40]);

[William Hanson]

all_LTG = [LTGlon,LTGlat];%combine the lat/lon LTG data into one matrix

j=4;%k clusters selected from kpickplot

[LTGidx2,LTGC2,LTGsum2,LTGD2] = kmeans(all_LTG,j,'replicates',10,'display','final');

ptsymb = {'bs','r^','md','go','c+','y*','k.'}; %assign symbols and colors for different clusters

for i=1:j LTGclust = find(LTGidx2==i); plot(all_LTG(LTGclust,1),all_LTG(LTGclust,2),ptsymb{i}); hold on end

plot(LTGC2(:,1),LTGC2(:,2),'ko');

plot(LTGC2(:,1),LTGC2(:,2),'kx');

hold off

title('LTG: Cluster Plot')

xlabel('Longitude º ')

ylabel('Latitude º ')

xlim([-35 55]);

ylim([-40 40]);

[Image Analyst]

Can you attach the data we need to run this? Make it easy for people to help you after you read this link

Please read how to format your code so I won't have to do it again for you.

Answer 1

[William Hanson]

Lightning data in .csv attached