Hi,
From the previous email we are allowed to parallelize. I am still waiting on the answer to the limit of time for running and memory usage limit.
However, for parallelization in R, is there special way that it needs to be done? I had submitted my code for SC1 round 2 (assuming that it is open) and it is stuck on the following typical code segment for a really long time.
```
nCores = min(10,detectCores() - 1)
cl<-makeCluster(nCores,outfile = "runningCh1.txt", type="FORK")
registerDoParallel(cl)
```
Are there any special requirements to parallelize in the machines they are running in?
Thanks,,
Created by SA90 There is no policy on parallelization, as long as you stay within in the alloted time/memory/processing. An e-mail will be sent out hopefully this afternoon detailing these.
SA90, you should feel free to try running it again, but unfortunately helping teams to parallelize their code is beyond what we can help with. i think in the end the organizers are unlikely to require parallelization but rely on the old framework. because parallelization itself is enough to disqualify over half of the teams. And it looks with the old framework, every finished in 1 week. but if they do, i think it would be a huge advantage to experienced teams, they can definitely modify their code to finish in less than 5 minutes at the sacrifice of some level of performance drop.
@thomas.yu @andrewelamb
was there a decision?