Afterwards you can sample a posterior with emcee, check that the chains have converged, and report parameters with credible intervals and their correlation.
Afterwards you can say what the condition number of a matrix measures, and predict how many digits a linear solve can lose before you run it.
Afterwards you can put a 3D field on a PyVista grid, cut it with slices, a threshold, and isosurfaces, and render the view to a static image for a paper.
Afterwards you can turn a parameter sweep into a readable animation with matplotlib.animation, update artists instead of redrawing them, and keep the GIF small.
Afterwards you can draw a gridded field and station points on a map projection with Cartopy, add coastlines from shipped files, and pick a projection that fits.