The Precision Problem in Floating Point Arithmetic: Avoiding Unexpected Results with High-Precision Arithmetic