On Fri, 12 May 2023, GIULIO PALOMBA wrote: I disagree. The problem is not related to dummy variables _per se_; the problem comes from the possibility that the Z matrix may not be full-rank in a given subset. I've tried a quick fix, but unfortunately, it's not general enough. I'll be unable to get any more work done over the weekend; I'll see what I can do next week, unless somebody beats me to it ;) ------------------------------------------------------- Riccardo (Jack) Lucchetti Dipartimento di Scienze Economiche e Sociali (DiSES) Università Politecnica delle Marche (formerly known as Università di Ancona) r.lucchetti(a)univpm.it http://www2.econ.univpm.it/servizi/hpp/lucchetti -------------------------------------------------------