Solving linear system using numpy

Question

I try to write a function that solves a linear system, and I get wrong values (15.7 instead of 16). I try to understand what might cause this mistake (rather than using an e