imo the fact that you have an implementation that ...

# datascience

[kyonifer]

imo the fact that you have an implementation that takes double is an implementation detail of the numerical precision you're able to offer currently (quite good)