[SIESTA-L] Help: Why is the Carbon pseudo potential coming with siesta 2.0 onlycontains 2s and 2p?

[You Lin]

Hello, dear siesta developers:

I was wondering why the Carbon pseudo potential coming with siesta 2.0 
only contains 2s and 2p? I figured this out since I can't set DZP in the 
basis set for carbon. Is there any particular reason for this?


Thanks.


Physics explains everything!
  -- You Lin




You Lin

Department of Physics
University of South Florida
4202 East Fowler Avenue
Tampa, FL 33620


Tel: (813) 974-8233 [Office]
Fax: (813) 974-5813
Homepage: http://shell.cas.usf.edu/~ylin

Re: [SIESTA-L] KB projector

[Oleksandr Voznyy]

atom: Maximum radius for r*vlocal+2*Zval:   27.49097
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: Rc( 0)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big


All this tells that you have the problems with your pseudo generation 
(too big radii).


You should try to take the cutoffs as big as possible to make the 
pseudopotential smoother (for s and p orbitals) but at the same time it 
should not be too big for d and f (non-bound) orbitals, so that 
non-local parts of KB projectors should be as short as possible.


Show your input file for pseudo.

Re: [SIESTA-L] KB projector

[Nichols A. Romero]

If you post you .fdf, we might be better able to help you.

On 2/5/07, Saswata Bhattacharya <[EMAIL PROTECTED]> wrote:


dear frnds,

i have got a very peculiar problem..whenever i make a test run with a very
simple .fdf file and the required .psf files it gives lots of warnings as
follows

"comlocal: WARNING It might be a good idea to increase
comlocal: WARNING parameter ntbmax (in file atmparams.f)
comlocal: WARNING to at least ntbmax =610
comlocal: WARNING It might be a good idea to increase
comlocal: WARNING parameter ntbmax (in file atmparams.f)
comlocal: WARNING to at least ntbmax =552
atom: Maximum radius for r*vlocal+2*Zval:   27.49097
GHOST: No ghost state for L =  0
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 0)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
GHOST: No ghost state for L =  1
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 1)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
GHOST: WARNING: Ghost state for L =  2
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 2)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
KBgen: WARNING: Ghost states have been detected
KBgen: WARNING: Some parameter should be changed in the
KBgen: WARNING: pseudopotential generation procedure.
Stopping Program from Node:0
Aborted"

but the thing is that no where i have written anything in the .fdf file
regarding user basis or KB projector as i made a very simple run with only
co-ordinates and some essential thingsi have seen in the manual that
neither user basis nor KB projector are essesntial to specify...so as these
are optional i didnt use these for my test runcan anybody give me any
suggestion?
with regards,
Saswata

--
Here's a new way to find what you're looking for - Yahoo! 
Answers





--
Nichols A. Romero, Ph.D.
1613 Denise Dr. Apt. D
Forest Hill, MD 21050
443-567-8328 (C)
410-306-0709 (O)

[SIESTA-L] KB projector

[Saswata Bhattacharya]

dear frnds,

i have got a very peculiar problem..whenever i make a test run with a very 
simple .fdf file and the required .psf files it gives lots of warnings as 
follows

"comlocal: WARNING It might be a good idea to increase
comlocal: WARNING parameter ntbmax (in file atmparams.f) 
comlocal: WARNING to at least ntbmax =610
comlocal: WARNING It might be a good idea to increase
comlocal: WARNING parameter ntbmax (in file atmparams.f) 
comlocal: WARNING to at least ntbmax =552
atom: Maximum radius for r*vlocal+2*Zval:   27.49097
GHOST: No ghost state for L =  0
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 0)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
GHOST: No ghost state for L =  1
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 1)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
GHOST: WARNING: Ghost state for L =  2
KBproj: WARNING: KB projector does not decay to zero
KBproj: WARNING: parameter Rmax in routine KBgen should be increased
KBproj: WARNING: Rc( 2)=  6.1334
KBproj: WARNING: Cut of radius for the KB projector too big
KBproj: WARNING: Increasing the tolerance parameter eps
KBproj: WARNING: might be a good idea
KBgen: WARNING: Ghost states have been detected
KBgen: WARNING: Some parameter should be changed in the 
KBgen: WARNING: pseudopotential generation procedure.
Stopping Program from Node:0
Aborted"

but the thing is that no where i have written anything in the .fdf file 
regarding user basis or KB projector as i made a very simple run with only 
co-ordinates and some essential thingsi have seen in the manual that 
neither user basis nor KB projector are essesntial to specify...so as these are 
optional i didnt use these for my test runcan anybody give me any 
suggestion?
with regards,
Saswata 


-
 Here’s a new way to find what you're looking for - Yahoo! Answers 

Re: [SIESTA-L] Bug in bands.F ?

[Vasilii Artyukhov]

Sorry, could you please explain this part in some more detail?

2007/2/2, Fabio Bernardini <[EMAIL PROTECTED]>:


Hello

In version 2.0 it seems there is a bug in Bands.F file.
When you run a spin polarized bandstructure
the values for the fermi energy in the first line of .bands
are wrong.
I think that the problem is in bands.F

   475 C Write Fermi energy
   476 if(nspin .eq. 2) thenWRONG ?
   477  write(iu,*) efs(1)/eV,efs(2)/eV
   478 else
   479  write(iu,*) ef/eV
   480 endif

I fixed the problem by in the following way

   475 C Write Fermi energy
   476 if((nspin .eq. 2).and.(fixspin)) then
   477  write(iu,*) efs(1)/eV,efs(2)/eV
   478 else
   479  write(iu,*) ef/eV
   480 endif

Fabio

Re: [SIESTA-L] User.Basis

[Oleksandr Loboda]

Thank you for the reply. But your message does not give clear answer on my
question. If you insist that it is possible then would you be so kind to
show us HOW one can use block PAO.UserBasis in order to set up manually  for
example  the following basis set into siesta:

 C   0
 S   9  1.00
   33980. 0.91000E-04
   5089.0 0.70400E-03
   1157.0 0.36930E-02
   326.60 0.15360E-01
   106.10 0.52929E-01
   38.110 0.147043000
   14.750 0.305631000
   6.0350 0.399345000
   2.5300 0.217051000
 S   9  1.00
   33980.-0.19000E-04
   5089.0-0.15100E-03
   1157.0-0.78500E-03
   326.60-0.33240E-02
   106.10-0.11512E-01
   38.110-0.34160E-01
   14.750-0.77173E-01
   6.0350-0.141493000
   2.5300-0.118019000
 S   1  1.00
  0.73550  1.
 S   1  1.00
  0.29050  1.
 S   1  1.00
  0.0  1.
 P   3  1.00
   34.510 0.53780E-02
   7.9150 0.36132E-01
   2.3680 0.142493000
 P   1  1.00
  0.81320  1.
 P   1  1.00
  0.28900  1.
 P   1  1.00
  0.10070  1.
 D   1  1.00
   1.8480  1.
 D   1  1.00
  0.64900  1.
 D   1  1.00
  0.22800  1.
 F   1  1.00
   1.4190  1.
 F   1  1.00
  0.48500  1.
 G   1  1.00
   1.0110  1.
 S   1  1.00
  0.41450E-01  1.
 P   1  1.00
  0.32180E-01  1.
 D   1  1.00
  0.76600E-01  1.
 F   1  1.00
  0.18700  1.
 G   1  1.00
  0.42400  1.

Thanks in advance

Re: [SIESTA-L] User.Basis

[Vasilii Artyukhov]

Perhaps you should consider consulting the literature about various basis
set kinds and their peculiarities: Gaussian functions, all sorts of
plane waves and wavelets, localized muffin-tin orbitals, numerical atomic
orbitals, finite grids and splines... -- just to understand why you
don't want Gaussian bases in SIESTA or why you can't do an all-electron
calculation with it.

The answers to your questions are: (1) SIESTA does support any basis
functions that are a product of a radial part and a spherical harmonic AND
have a finite support, and hence, (2) it IS possible to set them manually in
input (almost exactly), but it's like the last thing anybody would want to
do with bases in SIESTA. If you need both linear scaling and Gaussian basis
sets, your choice is pretty much limited to Gaussian 03 (which is, to my
opinion, not quite good) and CP2k (they use a combined Gaussian--plane wave
basis set for KS orbitals and density, respectively).

2007/2/5, Oleksandr Loboda <[EMAIL PROTECTED]>:


If I understood you correctly SIESTA does not support any of Dunning-Hay,
Huzinaga basis sets? And it is impossible to set them manually in input.

Re: [SIESTA-L] User.Basis

[Oleksandr Loboda]

If I understood you correctly SIESTA does not support any of Dunning-Hay,
Huzinaga basis sets? And it is impossible to set them manually in input.