Python: floating point precision in matrices

Question

following problem: I have a transition Matrix from a graph. I assigned each edge of a given vertex a transition-probability p: 1/number_outedges.

Natural