1: Reducing Randomness via Irrational Numbers 2: Linear-Time Approximation Algorithms for Computing Numerical Summation with Provably Small Errors