[Thermo_pw-forum] a few questions regarding thermo_pw

[Andrea Dal Corso]

Quoting Suza W <suza.rri at gmail.com>:

> Dear All,
>
> I am having following questions regarding thermo_pw :
>
>  (1)
>
>  plot  "output_pband.dat.g1.1.1" u ($1*xscale-xshift):($2*fact-eref)*gfact
> w l lw band_lw lc rgb color_red, \
>  "output_pband.dat.g1.1.2" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_green, \
>  "output_pband.dat.g1.1.3" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_blue, \
>  "output_pband.dat.g1.1.4" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_cyan, \
>  "output_pband.dat.g1.2" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.3.1" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.3.2" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_green, \
>  "output_pband.dat.g1.3.3" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_blue, \
>  "output_pband.dat.g1.3.4" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_cyan, \
>  "output_pband.dat.g1.4.1" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.4.3" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_blue, \
>  "output_pband.dat.g1.4.5" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_magenta, \
>  "output_pband.dat.g1.5" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.6" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.7" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.8" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>  "output_pband.dat.g1.9" u ($1*xscale-xshift):($2*fact-eref)*gfact w l lw
> band_lw lc rgb color_red, \
>
>
> What is the underlying physics for using the same color (for example,
> color_red)  for all  the g1.n.1  outputs ?

Color codes are explained in the Doc/point_groups.pdf file.
They indicate the irreducible representations of a given
band or phonon dispersion. Along some lines (at zone border)
the symmetry analysis is not available and all bands have the
same color (red). If you are not in one of these two cases,
please provide more info about your input.


>
> (2) Why I am getting the following error even when max_geometries=1000000 ?
>
> The code stops because max_geometries is****

Max geometries is a new feature. This is probably a bug.
Thank you for telling me.

>
>
> (3) I was checking all the geometries and I found that
> a few phonons are imaginary only for geometries, too much expanded.
> Does there exist any possibilty presently in the thermo_pw code, which can
> restrict geometries with a lower expansion to get rid of the imaginary
> phonon issue ?
>

No. Presently the user must choose the number of geometries or the  
step between different geometries so that all phonon dispersions are  
positive at all geometries. See the user's
guide at what='mur_lc' to find how to specify geometries.

HTH,

Andrea


> Thanks and regards,
> Suza W.
> _______________________________________________
> Thermo_pw-forum mailing list
> Thermo_pw-forum at qeforge.qe-forge.org
> http://qeforge.qe-forge.org/mailman/listinfo/thermo_pw-forum