Re: [Open Babel] Bug in vibration code?

[David L. <lon...@gm...>]

On Sun, Mar 22, 2009 at 5:08 PM, David Lonie <lon...@gm...> wrote:
> I've sent this to both lists, since there seems to be a bug in openbabel, but I noticed and can replicate it in avogadro:
>
> In libavogadro/src/extensions/vibrationdialog.cpp, add some debug output after the frequencies and intensities are loaded in setMolecule, e.g:
>
>    // OK, we have valid vibrations, so add them to the table
>    vector<double> frequencies = m_vibrations->GetFrequencies();
>    vector<double> intensities = m_vibrations->GetIntensities();
>    qDebug() << "size intensities   " << intensities.size();
>    qDebug() << "size frequencies   " << frequencies.size();
>
> Recompile and open methane.out in testfiles, then open the vibration extension. Output shows:
>
> size intensities    12
> size frequencies    9
>
> Shouldn't these be the same?
>
> Dave
>

I did a bit more digging, and here are the frequencies:

0   1575.73
1   1581.4
2   1584.19
3   1718.85
4   1723.36
5   2710.59
6   2770.6
7   2775.4
8   2778.94

And the intensities:

0   0
1   13.763
2   12.309
3   12.408
4   0
5   0.005
6   0.005
7   0.009
8   0
9   51.922
10   48.934
11   48.648

Those zeros shouldn't be there.  In the methane.out file, the
vibrational mode data is arranged in three groups of three, so it
looks like there is an extra zero being thrown in when each row is
being read. This is using the 2.2.x branch of svn openbabel and trunk
avogadro.

Please cc me if replying on openbabel-discuss, I've not actually joined it ;)

Dave