Localization in a MD simulation on open-shell systems: Wannier centers for spin "1" MOs not printed on completion of the first localization process

[Max]

Dear CP2K users and developers, 


In restarting a MD simulation on an open-shell system, I've asked for the 
localization of the occupied MOs. 

      &LOCALIZE
        EPS_LOCALIZATION 1.0E-4
        METHOD CRAZY
        MAX_ITER 1000
        USE_HISTORY .TRUE.
        &PRINT
          &WANNIER_CENTERS
            IONS+CENTERS .TRUE.
            FILENAME =wannier.xyz
            &EACH 
              MD 10
            &END EACH 
          &END WANNIER_CENTERS
        &END PRINT
      &END LOCALIZE

The calculations go fine but, (i) at the end of the first localization 
process, 
only the Wannier centers for the spin-down ("2") MOs are printed; and (ii) 
at the end of the subsequent ones, the centers are printed for both types 
of MOs. This happens whenever I restart the simulation. Could this be a 
bug in the first call to the print routine? Or am I doing something wrong? 
I'm running c2pk-2.6.0. 


Thanks in advance for your help

Regards,
Max
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150320/6eb91f1b/attachment.htm>