Using sgkit for population genetics analysis with xarray

Sgkit is the most advanced Python library for doing population genetics analysis. It’s a modern implementation, leveraging almost all of the fundamental data science libraries in Python. When I say almost all, I am not exaggerating; it uses NumPy, pandas, xarray, Zarr, and Dask. NumPy and pandas were introduced in Chapter 2. Here, we will introduce xarray as the main data container for sgkit. Because I feel that I cannot ask you to get to know data engineering libraries to an extreme level, I will gloss over the Dask part (mostly by treating Dask structures as equivalent NumPy structures). You can find more advanced details about out-of-memory Dask data structures in Chapter 11.

Getting ready

You will need to run the previous recipe because its output is required for this one: we will be using one of the PLINK datasets. You will need to install sgkit.

As usual, this is available in the Chapter06/Sgkit.py Notebook...