Given a collection of members, we want to find the two that have the greatest distance from each other in some dimension.
Given a collection of members, we want to find the two that have the greatest distance from each other in some dimension. This is the "linear" strategy.
Other strategies (like finding the greatest distance in both dimensions) might give better seeds but would be slower. This seems to work OK for now.
Splits the children of a branch node.
Splits the children of a branch node.
See splitter for more information.
Splits the children of a leaf node.
Splits the children of a leaf node.
See splitter for more information.
Splits a collection of members into two new collections, grouped according to the rtree algorithm.
Splits a collection of members into two new collections, grouped according to the rtree algorithm.
The results (a vector and a bounding box) will be used to create new nodes.
The goal is to minimize the area and overlap of the pairs' bounding boxes. We are using a linear seeding strategy since it is simple and has worked well for us in the past.