Clustering of lines

Roe asked 8 months ago

Hi. I loved the piece you did on HDBSCAN. I have a problem I would love to solve using the algorithm. I have a group of lines defined by their start points (x1, y1, z1) and end points (x2, y2, z2). The lines may be in hundreds or thousands. I want to find a region(s) in 3D space in which the lines are densest and then proceed to find the point through which most of them pass or almost pass. How do I implement this in Python?

3 Answers
Best Answer
Chris Staff answered 8 months ago

I am not sure if DBSCAN is the algorithm you will need. From what I understand, you are trying to find an intersection point or a point that is close to an intersection point for all the lines. This answer may be what you need: https://stackoverflow.com/a/52089867

Roe answered 8 months ago

Hi, Chris. Thanks for your response. Yes, I am trying to find the point that is close to intersection of lines. The link has been valuable. I have managed to find the point using scipy.minimize function. However, I would love to know more about DBSCAN and other clustering algorithms. I will need to know about them for my project. Besides you webpages do you have notes that you can share or recommend? By the way, keep up the good work!

Chris Staff replied 8 months ago

Hi Roe, thanks for your response! My articles are my notes, really — I write down everything and then convert it into an article 🙂 I have covered a few clustering algorithms here: https://www.machinecurve.com/index.php/how-to-use-scikit-learn-for-machine-learning-with-python-mastering-scikit/#clustering

Roe answered 8 months ago

Chris that is brilliant. I am sure to follow your articles. Your work is really great. You distill the complex concepts into simpler but still interesting chunks. Thanks again.

Your Answer

12 + 8 =