[yt-users] Nearest neighbours

[Elizabeth Tasker]

Hey Stephen,

How complex would that be? Currently, I've cut out an extracted region and I'm working around the neighbours of cells to decide if they deserve to stay in what I'm defining as a "cloud". It's basically fine, I just like the idea of working directly with the raw data because it seems the less steps there are between raw and the result, the less numerical errors there are likely to be. That said, including different sized grid cells might make my potential field bumpy so it might not necessarily work!

It also seems to me that having a routine to find the neighbours of a grid cell, independent of whether they are on the same grid, would be a useful general analysis tool. 

So ... how hard are we talking? :)

Elizabeth


On 2011-09-12, at 9:28 PM, Stephen Skory wrote:

> Elizabeth,
> 
>> Is there a routine in yt for returning the nearest neighbours to a cell? I
>> was looking at Sam's amr_kdtree and it mentions bricks.... what's a brick?
>> Is a cell a brick? Is a brick a cell? Is that a deep question?
> 
> Sam's stuff wouldn't be appropriate for what I'm guessing you want,
> but you could probably use the same kdtree that the halo finder uses.
> Let me know if you're still interested.
> 
> -- 
> Stephen Skory
> s at skory.us
> http://stephenskory.com/
> 510.621.3687 (google voice)
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org