[Nek5000-users] trouble with visit

Wed Mar 20 10:49:49 CDT 2013

Hi Joerg,

The first " 4 " indicates that the data is output in 4-byte (32-bit)
format.

My guess is that there might be a problem with the raw data
in the fields that are being written.  For example, what is
the sign of epst() ?  Is it guaranteed to be positive ?

       do i=1,n
          epst(i)=log(epst(i)+tiny)
          epsv(i)=log(epsv(i)+tiny)
       enddo

Paul

On Wed, 20 Mar 2013, nek5000-users at lists.mcs.anl.gov wrote:

> Hi Paul,

I didn't understand what Justin meant.
The first line in all output files looks like this:

#std 4 12 12 12      64000     256000   .2165378385933E+03      5000
  2      4 XU

Thanks Joerg.

> Hi Joerg,
> 
> Did you try Justin's suggestion of looking at your header?
> It probably already is 32-bit ? Â  (This would normally be
> the case.)
> 
> Perhaps the output data is corrupted, e.g., nans or whatever...?
> 
> Paul
> 
> 
> ----- Original Message -----
> From: nek5000-users at lists.mcs.anl.gov
> To: nek5000-users at lists.mcs.anl.gov
> Sent: Wednesday, March 20, 2013 10:03:10 AM
> Subject: Re: [Nek5000-users] trouble with visit
> 
> p63 is set to zero and it still does not work
> 
> Joerg
> 
> Am 20.03.2013 um 14:47 schrieb nek5000-users at lists.mcs.anl.gov:
> 
>>
>> It looks like setting p63=0 should take care of that...
>>
>> Paul
>>
>>
>> On Wed, 20 Mar 2013, nek5000-users at lists.mcs.anl.gov wrote:
>>
>>> Hi Paul,
>>
>> how can I set up the single precision output?
>>
>> best regards, Joerg.
>>
>>> Hi Joerg,
>>> Normally the output is indeed in single precision.
>>> It's not clear to me the double-precision + visit
>>> is working - but I've been a bit out of the loop
>>> on this.
>>> Paul
>>> On Wed, 20 Mar 2013, nek5000-users at lists.mcs.anl.gov wrote:
>>>> Hi Neks,
>>>>
>>>> I am facing problems with the visualization in visit.
>>>> The output of the three scalar fields uzt, epst, epsv
>>>> is done is seen in the following line from my usr file.
>>>>
>>>> Here, I wrote the data into four files (see further below).
>>>> I also attach the visit Â script
>>>>
>>>> !-------usr file-------------------------------------------------
>>>> Â  Â  Â  tiny=1.e-16
>>>> Â  Â  Â  do i=1,n
>>>> Â  Â  Â  Â epst(i)=log(epst(i)+tiny)
>>>> Â  Â  Â  Â epsv(i)=log(epsv(i)+tiny)
>>>> Â  Â  Â  enddo
>>>>
>>>> !-------No pressure field, no scalar output
>>>> Â  Â  Â  ifpo=.false.
>>>> Â  Â  Â  ifto=.false.
>>>>
>>>> Â  Â  Â  call outpost2( uzt,epst,epsv,pr,t,0,'eps')
>>>>
>>>> Â  Â  Â  ifpo=.true.
>>>> Â  Â  Â  ifto=.true.
>>>>
>>>>
>>>>
>>>> !-------Visit script------------------------------------------------
>>>> NEK5000
>>>> version: 1.0
>>>> filetemplate: epsG1_64%d.f%05d
>>>> firsttimestep: 1
>>>> numtimesteps: 1
>>>>
>>>>
>>>>
>>>> !-------the four files that have to be load into visit and which make up one snapshot--------------
>>>> -rw-r--r-- 1 josc1234 linux 2657536136 27. Feb 08:03 epsG1_640.f00001
>>>> -rw-r--r-- 1 josc1234 linux 2657536136 27. Feb 08:09 epsG1_641.f00001
>>>> -rw-r--r-- 1 josc1234 linux 2657536136 27. Feb 08:16 epsG1_642.f00001
>>>> -rw-r--r-- 1 josc1234 linux 2657536136 27. Feb 08:22 epsG1_643.f00001
>>>> -rw-r----- 1 josc1234 linux Â  Â  Â  Â  87 20. MÃ¤r 13:37 vis.nek5000
>>>>
>>>>
>>>>
>>>> !-------This is the error message in visit---------------------------------------
>>>> The compute engine running on makalu51.rz.tu-ilmenau.de has exited abnormally.
>>>>
>>>> Shortly thereafter, the following occured...
>>>>
>>>> The Pseudocolor plot of "y_velocity" for the file "/usr/wrk/people4/josc1234/nek5_visit/r095_Lx1_12_Ra1e8/vis.nek5000" could not be generated by the compute engine on host "localhost".
>>>>
>>>>
>>>>
>>>> My questions:
>>>>
>>>> Could it be that the graphics output should be in single precision?
>>>> I saw that the people at KTH did have similar trouble.
>>>> If yes, how can I switch to single precision in the usr file output?
>>>> For smaller grid it worked.
>>>>
>>>> Thanks in advance, Joerg.
>>>>
>>>>
>>>> _______________________________________________
>>>> Nek5000-users mailing list
>>>> Nek5000-users at lists.mcs.anl.gov
>>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>>>>
>>
>> _______________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.gov
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users_______________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.gov
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> 
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users