[petsc-dev] checking nightly build examples?

[Matthew Knepley]

On Mon, Dec 2, 2013 at 7:01 AM, Barry Smith <bsmith at mcs.anl.gov> wrote:

>
> On Dec 2, 2013, at 6:15 AM, Matthew Knepley <knepley at gmail.com> wrote:
>
> > On Sun, Dec 1, 2013 at 10:36 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:
> >
> >    Since the nightly examples are difficult to check even when correct
> due to floating point differences how about simply removing the floating
> point numbers before diffing? I did this on next and got
> >
> > I pushed a simple system for checking SNES and KSP monitor with a
> numerical diff. It is running now
> > on SNES ex62 and doing well.
>
>    Where is it and how does one run it?


It is a set of simple classes in builder.py starting at line 426. Here is
how it works:

  1) All solvers must have both *_monitor_* and *_converged_reason

  2) We create a parser, like SolverParser or IdentityParser, and run it
over the output

You can see the usage starting at line 1614 of builder.py. It is very very
simple, but I did
not think it was worth doing anything complex.

    Matt


> >
> >    Matt
> >
> > 1860 alltests.log
> >      130 alltests.log.short
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > --
> > What most experimenters take for granted before they begin their
> experiments is infinitely more interesting than any results to which their
> experiments lead.
> > -- Norbert Wiener
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20131202/955a03cb/attachment.html>