This seems like overkill? But for really large samples it could be useful.
On the other hand, you could let people do that for themselves. pval + nboots tells us exactly how many it was bigger than, and we could run that same bit of code a hundred times across cores, and calculate the resulting pval easily.
Seems like I could just explain that inside the manual. Lets not add a dependency on parallel or something ridiculous.
This seems like overkill? But for really large samples it could be useful.
On the other hand, you could let people do that for themselves. pval + nboots tells us exactly how many it was bigger than, and we could run that same bit of code a hundred times across cores, and calculate the resulting pval easily.
Seems like I could just explain that inside the manual. Lets not add a dependency on parallel or something ridiculous.