Isolate and group regions on a 3D plot based on sensitivy to a 4th parameter

I've got a reasonably large dataset roughyl 300,000x4. The first 3 columns relate to the x,y,z coordinates of a component in space and the 4th column is a rating based on certain factors and characteristics of the component at that x,y,z coordinate (i.e. temp, stress, strain, material, safety factors etc.) . Some example of the dataset rows are below:

X Y Z Rating

11.67 -5.70 341.21 24860692

-12.39 6.30 389.70 14831798

8.90 -3.12 368.10 17861810

13.80 -12.41 355.95 14962000

I want to identify the regions of the component which have a low rating and are at risk. I've found the individual nodes simply using the find function as below:

w = find(Rating <= 1E5);
u(:, 1) = X(w);
u(:, 2) = Y(w);
u(:, 3) = Z(w);
u(:, 4) = Rating(w);

When I plot this, it produces the following:

It is clear there are several different regions (about 9 or 10 maybe) of the component which have a low rating. I want to identify these grouped regions and output the x,y,z coordinates which enclose each of the regions. I'm guessing I somehow need to identify which x,y,z points are close together within the smaller dataset of low rating points but I'm not sure how to do this as the regions are not a set size and will vary when I run this code for different components so I need something generic.

Answer 1

[darova]

Look

[Vicki Corrigan]

Thank you! This is exactly the sort of thing I was after :)

[darova]

Accept the answer. They won't pay me

[darova]

Try clusterdata() also