[Wien] Problems with Optimize (V, c/a)

Stefaan Cottenier Stefaan.Cottenier at fys.kuleuven.ac.be
Sat Nov 8 00:26:55 CET 2003


> Commandline: x optimize -up 

'optimize' does not need options (check by looking at the bottom line 
of the output from 'x optimize -h'), so your option '-up' will be 
neglected.

> Program input is: "1 1 -10,-5,0,+5,+10% " 

The first '1' means you select volume optimization. The second '1' 
that you want 1 structure change. Then follow 5 instead of 1 structure 
changes, which is not valid (and probably also the %-sign is illegal).

What causes your error:

> 'INILPW' - can't open unit: 18             
>  'INILPW' -        filename: FeAl111r2.89dum.vsp                     
>  'INILPW' -          status: old          form: formatted            
>  'LAPW1' - INILPW aborted unsuccessfully.

is probably this: your ran your FeAl as a magnetic calculation in this 
directory (runsp). Then you must run also optimize.job magnetically. 
The output tells you that *you* must make changes to the script to 
make it valid:

>  #   Modify this script according to your needs: 
>  #      Uncomment one of the lines ...
>  #      Change run_lapw to runsp_lapw or use different convergence
>         criterium

You did not do that, you left the original run_lapw instead of 
runsp_lapw :

>      run_lapw -ec 0.0001

Conclusion: go through optimize.job line by line, and make sure that 
it executes for every case.struct the kind of scf-cycle that you want.

Stefaan




More information about the Wien mailing list