[Wien] Fundamental question on constraint dft calculation of effective U by Madsen and Novak

For core states we do some "naive" approximation (which is usually ok): Although we have a fully relativistic description with j as quantum numbers (and as you say, s is NOT a good q.n.), we neglect this fact and perform two independent calculations, one in the spin-up and the other one in the spi

[Wien] Fundamental question on constraint dft calculation of effective U by Madsen and Novak

Dear wien2k(ers), I'm using method suggest by Madsen and Novak (MN) (http://www.wien2k.at/reg_user/textbooks/Constraint_U.pdf) to calculate the effective U. I reproduced their results. However I have some fundamental (and maybe naive) question: 1. In the Anisimov paper (cited by MN), they use m,