An adaptation of Similarity Network Fusion (SNF) to Python
While an implementation of SNF in Python already exists (https://github.com/rmarkello/snfpy), I found that the output from their main SNF function differed from the output I got from using the SNF package developed by the original authors in R (see tests/test_snf.py in this repository).
snfpy is a great library in its own right (some of its functions are
used here as a dependency). I built this small package as an adjunct to provide an
SNF function that is an exact replica of the widely used SNFtools package in R.
This pacakge also includes an implementation of the robust_core_clustering
function from Jacobs et al. (2021).
After cloning this repository locally, you can install it by navigating to the main package folder and running the command
pip install .
To import the main SNF function, you can use this line of Python code:
from snftools.snf import SNF