Hacker News new | ask | show | jobs
by dvko 896 days ago
I tried the linear search by station name in my first naive approach. Using a hashmap was at least 2-3x as fast with the ~415 distinct keys in the 1BRC dataset.