The ABINIT Users Mailing List ( CLOSED )

[Christopher Taylor <cdtaylor@lanl.gov>]

Marc & Natalie,

I'm using the latest version on the website.

The READ problem seems to come from the reading of the INITOCC block 
in the U.atomicdata file.  The 'integer' value in the atom.atomicdata 
file is written as if a floating point number (i.e. instead of '2', 
it would write '2.00000'.)  *If I manually change the occupations to 
integer form, the code runs without the read errors.*  This must be 
something specific for my system (Mac OS X 10.4 on G5).

Thanks for your help,

Chris

> Hi Christopher,
>
> > I'm trying to build a PAW for Uranium.  I think I have finally got
> > atompaw to do what I want (after staring at the input file-structure
> > for many hours), but when I run the atompaw2abinit I get the
> > following text:
> >
> > "A READ statement using decimal base input found the invalid digit
> > '.' in the input file.  The program will recover by assuming a zero
> > in its place."
>
> You perhaps have a wrong version of atom2abinit tool. The last 
> official one is given on atompaw web site... but a new is coming 
> soon...
> But the error doesn't seem to be a fatal one.
>
> > although it apparently runs with no fatal errors.  Is this READ message
> > something I need to worry about?
> >
> > I've tried many types of PAW (with SR, no-SR, with loggrid, without
> > loggrid, Vanderbilt, VNCT, etc.), and atom2abinit always gives the
> > same result.  When I try to use the resulting PAW(s) in abinis, I get
> > funny errors which I can't find any info on in the forum (can't find
> > fermi level on 60 bissections, and/or vtorho errors like "The minimal
> > occupation factor is  1.900 ... should be at most 0.0", as well as 
> > "mkdenpos :
> > Density went < 0" etc), so I am thinking that maybe
> > the PAW conversion is just not right somewhere along the line.  At
> > least, that's where I'd like to start my troubleshooting and work my
> > way down the line to input parameters, ecuts, etc.
> Your projectors are too "high".
> Such errors occur when the "Rhoij" terms computed by Abinit are too 
> big. This is probably the case here. This signs that your projectors 
> are not suitable for a good Abinit calculation. Try other parameters 
> for Atompaw  (change reference energies, radii, ...); changing 
> projector scheme or grid is not the solution... but changing 
> energies and radii is..
>
> Marc Torrent


--
-- Christopher Taylor, Ph.D.
Materials Science and Technology, MST-6
MS G755
Los Alamos National Laboratory
Los Alamos, New Mexico 87545

v: 505-667-5499
f: 505-667-2264