Ripley's K Function (Second order intensity function) Python

Question

I am looking for Ripley's k function implementation in Python. But so far haven't been able to find any spatial modules implementing this in scipy or elsewhere.

I have created Voronoi tessellation of a fibre composite and need to perform analysis using Ripley's K and pair distribution functions compared to a Poisson distribution.

Cannot upload images-not enough rep.

Catalano

Would it be an option for you to simply call an R function via RPy? In that case you can use the spatstat package for R. – Veto

Thanks! Didn't know about RPy before but that sounds feasible. – Catalano

Answer 1

Solved my problem, this is for others looking to do same analysis.

Definitely recommend using R for spatial analysis. A transfer from python is simple because all you need is coordinates of your point pattern.

1. Write a csv of x,y and z coordinates of your points using python

2. R has good functionality of reading csv using command read.csv("filename"). Make sure the directory is set properly using setwd command.

3. Convert the csv you just read to a ppp(point pattern which R understands) using as.ppp command.
4. Continue to use Kest, Gest etc for required spatial analysis.

Cheers.

Catalano

Answer 2

Now (2021) there is a Ripley's K implementation for Python, easy to install

https://pypi.org/project/ripleyk/

And also a more comprehensive package that includes Ripley's K, G, F and includes MC simulations for p-values, very nice.

https://github.com/pysal/pointpats/blob/master/notebooks/distance_statistics-numpy-oriented.ipynb

Hakodate