[petsc-dev] -ksp_monitor and different norms

Jed Brown jedbrown at mcs.anl.gov
Tue Oct 30 16:26:03 CDT 2012


Cool, looks good to me.

On Tue, Oct 30, 2012 at 2:24 PM, Mark F. Adams <mark.adams at columbia.edu>wrote:

> -ksp_monitor_max
> -ksp_norm_type unpreconditioned
> -ksp_monitor_short
>
> On Oct 30, 2012, at 5:18 PM, Jed Brown <jedbrown at mcs.anl.gov> wrote:
>
> On Tue, Oct 30, 2012 at 2:13 PM, Mark F. Adams <mark.adams at columbia.edu>wrote:
>
>> I have it set up now to print:
>>
>> >   0 KSP true resid max norm 1.149095087619e-03 ||r(i)||/||b||
>> 1.000000000000e+00
>> >   0 KSP Residual norm 0.00424852
>>
>> Is this OK?
>>
>
> Was this run with -ksp_monitor_max -ksp_monitor?
>
>
>>
>> I would prefer (for my user) to not have the second line printed (with
>> the 2-norm) ...
>>
>
> yes, naturally
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20121030/efbef2d3/attachment.html>


More information about the petsc-dev mailing list