[Wien] DSTART fails for 'mini' trying to treat complex case as real

[Peter Blaha]

I'd expect you have a problem with your csh or awk

Please verify the script     x_lapw:

search for cmplx    in this file.

you shold find some lines with   "awk", where
it is checked automatically if you have inversion
or not.

I'd expect that this does not work on your system.

In the first line of x_lapw put:

#!/bin/csh -fx       or

#!/bin/tcsh -fx

to get debugging output, or use the tcsh, which may work.



Oleg Rubel schrieb:
> Dear Wien2k Users,
> 
> I am running WIEN2k_08.1 (Release 14/12/2007) in a 'mini' mode using the 
> command
> 
>      min -i 100 -s 10 -j 'run_lapw -p -I -i 40 -fc 0.5 -ec 0.0001 -cc 0.001'
> 
> In the beginning, I saw a strange error
> 
>      DSTART - Error
> 
> although dstart runs fine during the initialization (init_lapw). The 
> error-file says
> 
>      marc-hn:~/wien_work/GaAsBeta2_2x4> cat dstart.error
>      'DSTART' - can't open unit: 15
>      'DSTART' -        filename: GaAsBeta2_2x4.in2
>      'DSTART' -          status: old          form: formatted
> 
> The structure I calculate is a GaAs surface, which does not have an 
> inversion symmetry. Therefore, Wien2k should search for the *.in2c file, 
> but it does not as you can see from the error message above.
> 
> The dstart.def file (see below) also seems to be constructed for the 
> 'real' and not for the 'complex' version, although the former dstart.def 
> generated after inicialization was 'complex' (not shown here).
> 
>      marc-hn:~/wien_work/GaAsBeta2_2x4> cat dstart.def
>      6,'GaAsBeta2_2x4.outputd','unknown','formatted',0
>      13,'GaAsBeta2_2x4.in0_std','unknown',    'formatted',0
>      14,'GaAsBeta2_2x4.in0','old',    'formatted',0
>      15,'GaAsBeta2_2x4.in2',   'old',    'formatted',0
>      17,'GaAsBeta2_2x4.in1',   'old',    'formatted',0
>      16,'GaAsBeta2_2x4.test',        'unknown','formatted',0
>      20,'GaAsBeta2_2x4.struct',      'old',    'formatted',0
>      81,'GaAsBeta2_2x4.rsp',    'old',    'formatted',0
>      51,'new_super.clmsum',    'unknown','formatted',0
> 
> According to the log-file below, Wien2k calls 'dstart -super', but '-c' 
> switch is missing. Later on 'lapw1' is called with the '-c' option as it 
> should be.
> 
>      marc-hn:~/wien_work/GaAsBeta2_2x4> cat :log
>      >   (init_lapw) options:
>      Wed Apr  2 17:05:17 CEST 2008> (x_lapw) nn -f GaAsBeta2_2x4
>      Wed Apr  2 17:05:19 CEST 2008> (x) nn
>      Wed Apr  2 17:05:27 CEST 2008> (x) sgroup
>      Wed Apr  2 17:05:34 CEST 2008> (x) symmetry
>      Wed Apr  2 17:05:38 CEST 2008> (x) lstart
>      Wed Apr  2 17:08:04 CEST 2008> (x) kgen
>      Wed Apr  2 17:09:16 CEST 2008> (x) kgen
>      Wed Apr  2 17:09:34 CEST 2008> (x) kgen
>      Wed Apr  2 17:09:47 CEST 2008> (x) dstart -c
>      >   (min) options: -i 100 -s 10 -j run_lapw -p -I -i 40 -fc 0.5 -ec 0.0001 -cc 0.001
>      GaAsBeta2_2x4.inM has been created automatically
>      Wed Apr  2 17:28:13 CEST 2008> (x) dstart -super
>      >   (min) recover inm-file & call job run_lapw -p -I -i 40 -fc 0.5 -ec 0.0001 -cc 0.001
>      >   (run_lapw) options: -p -I -i 40 -fc 0.5 -ec 0.0001 -cc 0.001
>      Wed Apr  2 17:28:13 CEST 2008> (x) lapw0 -p
>      Wed Apr  2 17:34:57 CEST 2008> (x) lapw1 -c -p
> 
> In principle 'mini' runs on the base of the density calculated in the 
> initialization, but it should not probably be like that.
> 
> I will be thankful for any pointers.
> 
> Oleg Rubel
> 
> ===========================
> Faculty of Physics
> Philipps University Marburg
> Germany
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien