Mixing geo location with other numerical values by using multidimensional points

Elastic Stack Elasticsearch
1

I read in this blog that geo location data, ie, longitude and latitude can be mixed with other numerical values as in the following graph from the blog.

"Note that the dimensions need not all be spatial! One exciting potential use case is mixing geo-spatial dimensions with other dimensional values, such as median household income, into a single 3D dimensional index. This would then allow fast range filters against both median household income and arbitrary geo-spatial shapes. Another example is indexing time as one dimension along with other numeric dimensions to support business-intelligence use cases."

It looks quite appealing to me, and I want to give it a try, But I don't know how to get started, specifically, what type to use when creating mappings and how to organize data when creating index?

2

I don't believe that has been exposed in ES as yet, the BKD tree stuff only just came in to ES 5.0.

3

Ok, I'll take a close look at the release notes to verify that, and if it's not possible yet, I will submit a feature request.

4

I didn't realize that your are working for elastic:joy:

5

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