Hello,
I'm working on noncollinear GGA+U+SOC calculations (using JUDFT_WARN_ONLY) for FeF3 in the fcc pyrochlore lattice using FLEUR. To generate a starting charge density, I begin with a collinear AFM configuration and run a self-consistent calculation to produce cdn. After that, I changed the magnetic configuration to an all-in/all-out (AIAO) noncollinear setup. To generate the AIAO input, I modify the .txt structure file accordingly and use:
inpgen -noco -nosym -overwrite -f inpFeF3.txt
Then I edit the generated inp.xml to assign the proper nocoParams for the AIAO spin directions. However, when I try to run FLEUR with this AIAO setup (using the previous AFM cdn1 files), I get the following error:

**************FLEUR-Error*****************
Error message: Number of atom groups in Fleur input file and in the charge density file don't match.
Error occurred in subroutine: loddop
Error from PE:0/4
*****************************************

I realize that this error occurs because the number of atom groups changes between the AFM and AIAO configurations, which makes sense. However, I'm unsure how to properly initialize the calculations so that I can switch to the noncollinear AIAO setup without encountering this issue

Thanks a lot. It seems it finally works. :)