Check this out, I am pretty much making something like this
https://github.com/simonccarter/sta
this project made me think of bigger datasets and speed and after some research and refactoring I was able to process 99,999,999 f64 values and make a few statistical calculations in 8.6842492s.
