Has anyone computed a Backwards lasso with positive coefficients? I would like the R commands or guidance on fitting such a model. The backwards lasso and positively constrained lasso are both described in Least Angle Regression, Efron etal 2002.