Parallel Calculations
Now, it's time to refactor the CalcTotalOrder function using parallel algorithms.
Arithmetic Operations in Parallel Algorithms
Another place where we can us parallel algorithms is CalcTotalOrder()
.
Instead of std::accumulate
we can use std::transform_reduce
.
As mentioned in the Parallel Algorithms chapter, the floating point sum operation is not associative. However, in our case, the results should be stable enough to give 2 decimal places of precision. If you need better accuracy and numerical stability, please consider using a different method.
Get hands-on with 1300+ tech skills courses.