sem.Rd
compute standard error of the mean
sem(x)
numeric vector, a sample of numbers
the standard error of the mean.
sem(1:5) #> [1] 0.7071068
