For many vine plots, I have NDVI and Leaf Area values for each vine. I already know that NDVI and LA has a strong positive correlation as you can see in this picture. enter image description here

But as you can see too, there are some points away from this pool of points. How can I cluster or identify these points? I'm using Python with sklearn, scipy and other libs like them.

enter image description here

  • 2
    $\begingroup$ This is called outlier detection. There are many ways to do it and the algorithm that performs best depends on the details of your case. $\endgroup$
    – Louic
    Jun 25 '19 at 13:33
  • $\begingroup$ Ok, thank you @louic, I'll try this! $\endgroup$
    – Tim C.
    Jun 25 '19 at 13:38

As already mentioned in a comment, there are some methods to detect outliers. The best known might be Cook‘s distance in (or after) regression. In essence you look for observations with a high residual.

Here is a short R tutorial. Here is a short Python approach.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.