Pandas dataframe round underlying data (not just how it's displayed)

Question

You may say this has been done before, but all solutions I can find fail, allow me to demonstrate:

first of all here are my imports

import numpy as np