Dear all,
I want to know what's the best index to measure the spatial clustering (aggregations) in 3D environment.
Thank you in advance
Ilhem
Depends on the type of spatial data that you are working with...
So...what type of spatial data do you have?
I'm working with data points that represents spatial positions of cells (each one has a spherical form).
OK. Do the cells represent different types? Or are you looking at just one type?
No just one type, and I'm looking for a index to test the aggregation formation (like Clark Evans index in 3D environment).
Sorry for the delayed response (I had to be in the field this morning).
Our software package (PASSaGE 2) can do a 3D K-function,
The spatstat package in R can also handle three dimensional point pattern objects (pp3) and you can do any of the major summary functions (K, G, F) for 3D data. In general, spatstat is a great package for point pattern analysis, the book is awesome, and the authors are super helpful if you hit a snag.
At the present time, I don't think you can do a global statistic (such as Clark Evans) for 3D data, but you're probably better off using a second order method anyhow.
The major limitation (at the present time) is that there is no software (that I am aware of) that can handle inhomogeneous point processes in three dimensions. So if you have inhomogeneous data (which is very common) you are stuck. You could still use something like a K-function as a starting point, but you many not be able to determine whether departure from the expected pattern is due to dependency or inhomogeneity.
Hope this helps!
.
- Badges
- Science topic