After some digging, it looks like the underlying C++ streams aren't
flushing. This means that python might not actually have all the
information when you print them out. We may have to enable a "flush"
function for these streams for better error reporting on the python side,
I'll need to investigate this a bit more.
Cheers,
Brian
On Wed, Jun 29, 2016 at 3:14 PM, DmitriR <xzf...@gmail.com> wrote:
> Brian - Sure. Attached:
>
> RDKit-test-warnings-01.ipynb
> dff.pkl
> screenshot1-warningsPrintToNotebook.pdf
> screenshot2-noWarningsPrint.pdf
>
> This has gotten stranger though. Now sometimes I get no visible output
> (screenshot 2).
>
> The total length of captured warnings still differs run to run, but now I
> noticed that it alternates *imprecisely*; see comment in screenshot2, cell
> 32). When I compared the sets of warnings produced on alternating runs
> (where the difference is substantial: 43k characters vs 38k characters),
> they are different because a large number of warnings do not get produced.
> I don't know what the smaller variations are due to.
>
> Python 3.5.1 :: Anaconda 2.4.0 (x86_64), OSX 10.11.5, jupyter 4.1.0,
> Firefox
>
> Thanks.
> Dmitri
>
>
> > On Jun 29, 2016, at 9:07 AM, Brian Kelley <fustiga...@gmail.com> wrote:
> >
> > Dmitri,
> > Could you send me the notebook that displays these issues? I can't
> reproduce them.
> >
> > Thanks,
> > Brian
> >
>
>
>
------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
