# sort on x and y of 2d points

This topic is 1863 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

## Recommended Posts

hello.
I would use a sort algorithm that works with multiple indices.
For example i would sort based on cx and cy a set of 2d point.
I think to x and y combined in a color and i wish extract the groups of 2d points that have a similar color.
Is the same thing as an ordering based on x and y?
thanks.

##### Share on other sites
It's unclear if you want sorting or clustering. For sorting, the algorithms don't change, but you have to provide a comparator function. The most common total order for 2D points is the lexicographical order: You first compare the x coordinates, and you only compare the y coordinates as a tiebreaker.

However, the result of sorting won't generally make a good [url="http://en.wikipedia.org/wiki/Cluster_analysis#Clustering_algorithms"]clustering algorithm[/url], which is what you seem to need.

##### Share on other sites
As Alvaro says above, you'll need to define some form of metric to compare the distance between two (x, y) "colors". The closest thing I can think of to your request is a kd-tree, which is basically a multidimensional binary tree, which sorts a list of points in k dimensions, which supports efficient nearest-neighbor search (find the N closest colors in the tree to some color, according to your definition of "closest") and range search (find all the colors in the tree closer to some color than some distance X, again according to your definition of "distance").

But it depends on what you want to do, you haven't given enough information. Edited by Bacterius