Are there any R packages available that let me directly impose a set of linear parameter restrictions when estimating the simple linear and/or GLS regression model and directly recover the restricted coefficients, predicted values and residuals? Can I do this via the use of contrasts under the model.matrix command? Thanks jatwood