（python） labeling points which are in the similar region of a 3-D coordinate system

Question

There are 2000+ points in a 3-D sapce (as picture 1) . The points are classified into 7 groups by K-Means methods. the points in picture 1 are presented like this:

point_ID_1    x        y        z       group_label
    1       0.254    0.025      0.007        1
    2       0.201    0.078      0.081        1
    3       0.106    0.153      0.238        2
    4       0.052    0.015      0.084        3
    5       0.078    0.005      0.159        7
   ...       ...      ...        ...         ...

there are other 5000+ points in the same 3-D sapce (as picture 2) . The (x,y,z) of points in the picture 2 is different from the points in the picture 2. But their distribution ranges are very similar.

the point in picture are presented as :

point_ID_2    x        y          z       
    1       0.354    0.025      0.028       
    2       0.125    0.128      0.324
    3       0.068    0.292      0.221
    4       0.051    0.086      0.128
    5       0.028    0.081      0.082
   ...       ...      ...        ... 

I want to label the points in the same region of picture 2 and picture 1 with the same label, as the picture 3 shows.

Noted that the points are not in the same surface but in a 3-D space

How can I do this? I searched the Internet for a long time but I didn’t find any solutions. Please help me. Thank you very much!

question from:https://stackoverflow.com/questions/66062856/python-labeling-points-which-are-in-the-similar-region-of-a-3-d-coordinate-sys

Answer 1

When you classified the first 2000 points in a group with Kmeans you obtained a classifier. You can use this classifier to decide for the other 5000 points in which group they belong. If you used the Kmeans from scipy, then you can use the function sklearn.cluster.KMeans.predict() for this.

You can find information about this in the documentation.