How do I get k nearest neighbors from a numerical vector based index?

Elastic Stack Elasticsearch
1

Say I indexed something like this:

{
{'l1':[1,2,3,4],'l2':[5,6,7,7]}
{'l1':[1,2,6,6],'l2':[4,3,2,7]}
...
..
{'l1':[1,2,3,24],'l2':[5,6,37,1]}

}

And I have a dictionary as {'l1':[2,3,4,5]}

and I want to find k values of 'l1' which are nearest in the vector space to it, how do I go about this? Is this possible to do in elasticsearch?

I'm using the python API

2

you may want to read about the new vector datatype that has been recently added.

In addition it might help, if your clear up your question a bit? Are the above separate documents? Is 2,3,4,5 your query vector?

A fully fledged example with sample documents might help to understand the setting a bit more.

3

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.