CODEWITHSUNDEEP
rmulticore
h.l.m
h.l.m

Reputation: 13485

speeding up as.POSIXct on windows

I know that you can speed up the as.POSIXct function by using pvec, from the multicore package, but it only works on unix systems such as a mac. Is there a Windows equivalent, that would generate comparable speeds?

Any benchmarking would be interesting to see...

Upvotes: 1

Views: 137

Answers (1)

Dirk is no longer here
Dirk is no longer here

Reputation: 368589

I fear that you are confusing two issues.

Nothing here "speeds up as.POSIXct", this remains an expensive operation. What pvec does is to run "chunks" of the vector in parallel through the conversion, which minimizes the number of function calls to as.POSIXct -- as you quote in your comment above.

pvec however relies on forking, and that is something Windows does not offer. There are some fundamental difference between the OSs that matter, this is one of them.

Upvotes: 1

Related Questions