r2.Rd
R^2
r2(y, mupred, log_weights)
vector of observed values (n)
matrix of posterior draws (S x n) of point predictions
matrix of standardized loo weights (S x n) on the log scale