In “Model selection and estimation in regression with grouped variables (2005)” , the authors (Yuan and Lin) claim that “It is easy to see that this is identical to the solution path of group LARS when X is orthonormal”. The piecewise linear solution path of group lasso is identical to the path of group lars if the design matrix X is orthonormal on page 56.
But I don’t have any idea how to check that claim. Why are the solution paths of the group Least Angle Regression Selection and the group LASSO identical to each other? Thanks!