Replying to @MattDowle
Confirming the finding: data.table::fread still wins. #rstats #rdatatable

Feb 12, 2021 路 6:40 AM UTC

3
11
1
93
I guess benchmarking is not as trivial as I always thought :) running on a Intel laptop:
1
1
When running on a laptop did you close other tasks that come and go in background, like Chrome especially? And what compiler flags? But that's fine detail. Not 3x is the point.
1
I'm thinking I'd show how irrelevant the difference is compared to how slow base::read.csv must surely be and then...
1
2
7
Is anyone surprised?