ELF for nspin=1 was wrong in v.6.8
The extension of ELF to spin-polarized cases had broken ELF in v.6.8. Not sure it is correct now (just less wrong) due to absence of tests
The extension of ELF to spin-polarized cases had broken ELF in v.6.8. Not sure it is correct now (just less wrong) due to absence of tests