Implements the v-statistic, a measure that compares the estimation accuracy of the ordinary least squares estimator against a random benchmark.
See the paper here: http://www.ncbi.nlm.nih.gov/m/pubmed/23661222/ which features an implementation in R.
I also wrote a python version, see here.
install with
Pkg.add("VStatistic")
run with
import VStatistic: vstat, sample_size
vstat(120, 4, .05) # get v-statistic
sample_size(3, .05, v=.9) # get needed (total n, n per predictor)