Older files (some of which contain example runs):
<ul>

    <li><a href="atompaw-4.1.1.0.tar.gz" target="_blank">
      atompaw-4.1.1.0.tar.gz</a> (5.5mb) 1/07/2021
      Francois Jollet and Marc Torrent implemented the LDA-1/2 formalism
      described in the publication --
       <a href="https://doi.org/10.1016/j.cpc.2019.107042" target"_blank">
      The ABINIT project: Impact, environment and recent developments </a>
      Additionally, solutions to the full Dirac equations  for all-electron
      wavefunctions and densities are now available and have been used
      by Nils Brouwer to evaluate spin-orbit and other effects in warm
      dense matter.



    <li> <a href="https://github.com/atompaw/atompaw">Mirror site </a> for AtomPAW package on
      github starting with version 4.1.0.6 thanks to Marc Torrent, who also updated the instructions      for installation on Mac OS.
    <li><a href="atompaw-4.1.0.6.tar.gz" target="_blank">
      atompaw-4.1.0.6.tar.gz</a> (5.5mb) 6/27/2019
      Francois Jollet and Marc Torrent implemented a test of the atomic
      pseudobasis functions for positive definite overlap matrix following
      a suggestion by   Dr.  Brecht Vestichel of Synopsys, Denmark.   The
      eigenvalues of the overlap matrix in the basis of the projector functions
      are evaluated and outputted to the (case) file.   If any of them are
      smaller than 10<sup>-10</sup>, the program stops with the suggestion
      that the pseudobasis is incomplete.     Additionally, core electron
      and the corresponding pseudo core kinetic energy densities are now
      calculated and outputted to the (case).xml file for possible use by
      ABINIT.



    <li><a href="atompaw-4.1.0.4.tar.gz" target="_blank">
      atompaw-4.1.0.4.tar.gz </a>(5.5mb) 9/2018  Introduced a check on the
        charge density sent to the exchange-correlation functionals. This was
        found to cause trouble when including the compensation charge
        in the exchange-correlation functional as is done in the current
        version of Quantum Espresso and in abinit running in the
        usexcnhat mode.  In these cases, when negative arguments of
        the exchange-correlation functional is detected, no data file is
        generated and the output gives the advice to change the
        magnitude of the pseudo core density function (usually
        accomplished by reducing the value of <i>r<sub>core</core></sub></i>).
        <a href='./mystery.pdf'>Link</a> to notes and
        <a href="https://doi.org/10.1016/j.cpc.2019.05.009">
        a short publication</a>) on this subject.

    <li><a href="atompaw-3.1.0.3.tar.gz" target="_blank">
       atompaw-3.1.0.3.tar.gz </a>(3.8mb) -  January 2014 -
       Older version of atompaw with
       contributions from Marc Torrent and Francois Jollet as well as several
       others.



    <li><a href="atompaw-4.1.0.5.tar.gz" target="_blank">
      atompaw-4.1.0.5.tar.gz</a> (5.5mb) 12/2018  Marc Torrent modified
       abinitinterface.F90 so that (case).abinit file correctly handles
       the case of pbesol, compatable with abinit using libxc.   Note that
       the (case).abinit datasets are superceded by the (case).xml files.



    <li><a href="atompaw-4.1.0.4.tar.gz" target="_blank">
      atompaw-4.1.0.4.tar.gz </a>(5.5mb) 9/2018  Introduced a check on the
        charge density sent to the exchange-correlation functionals. This was
        found to cause trouble when including the compensation charge
        in the exchange-correlation functional as is done in the current
        version of Quantum Espresso and in abinit running in the 
        usexcnhat mode.  In these cases, when negative arguments of 
        the exchange-correlation functional is detected, no data file is
        generated and the output gives the advice to change the
        magnitude of the pseudo core density function.
        <a href='./mystery.pdf'>Link</a> to notes on this subject.
      

    <li><a href="atompaw-4.1.0.3.tar.gz" target="_blank">
      atompaw-4.1.0.3.tar.gz </a>(5.5mb) 7/8/2018 Upgrade to the compatibility
      with libxc made by Marc Torrent.
      Thanks to Casey Brock (Vanderbilt U.)  logderiv.l files
      now also output the arctangents of the logderivatives, shifted by &pi; for
      continuity.  Casey improved his "phase unwrapping
      algorithm".
      In general, comparison of the all-electron and pseudo values of
      the shifted arctangents give a numerically smooth measure of the
      ability of the PAW dataset to represent the wavefunctions.
      Previously,
        Marc Torrent modified
      the code for compatibility with libxc versions &leq; 4,
      (although the code is not yet ready
      for META-GGA or hybrid functionals). The ./configure command
      has been simplified.   For example to compile with libxc, the
      following command can be used: <br />
      ./configure --enable-libxc --with-libxc-prefix="/path-to-libxc" <br />
      Similarly, a new prefix command has been added for the linear
      algebra libraries: <br />
      ./configure --with-linalg-prefix="/path-to-blas-lapack" <br />

    <li><a href="atompaw-4.1.0.2.tar.gz" target="_blank">
      atompaw-4.1.0.2.tar.gz </a>(5.4mb) 5/2018 Intermediate
      update of the code mostly for libxc 4.1.1. Some bugs in this version
      have been corrected in version 4.1.0.3.        <br />
 
    <li><a href="atompaw-4.1.0.1.tar.gz" target="_blank">
      atompaw-4.1.0.1.tar.gz </a>(5.4mb) 3/8/2018 Marc Torrent modified
      the code for compatibility with libxc v4.0.0, 
      although the code is not yet ready
      for META-GGA or hybrid functionals). The ./configure command
      has been simplified.   For example to compile with libxc, the
      following command can be used: <br />
      ./configure --enable-libxc --with-libxc-prefix="/path-to-libxc" <br />
      Similarly, a new prefix command has been added for the linear
      algebra libraries: <br />
      ./configure --with-linalg-prefix="/path-to-blas-lapack" <br />
      

   <li><a href="atompaw-4.0.1.0.tar.gz" target="_blank">
      atompaw-4.0.1.0.tar.gz </a>(5.4mb) 9/9/2017 Marc Torrent modified
      the code for compatibility with libxc v3.0.0 which has been tested
      with LDA and GGA functionals (although the code is not yet ready
      for META-GGA or hybrid functionals). The ./configure command
      has been simplified.   For example to compile with libxc, the
      following command can be used: <br />
      ./configure --enable-libxc --with-libxc-prefix="/path-to-libxc" <br />
      Similarly, a new prefix command has been added for the linear
      algebra libraries: <br />
      ./configure --with-linalg-prefix="/path-to-blas-lapack" <br />



    <li><a href="atompaw-4.0.0.14.tar.gz" target="_blank">
      atompaw-4.0.0.14.tar.gz </a>(5.4mb) 12/28/2016  NAWH corrected
      the FINITE-NUCLEUS option for DFT calculations, allowing for models
      2, 3, 4, and 5 described by Andrae in 
      <a href="http://doi.org/10.1016/S0370-1573(00)00007-7">
      Physics Reports <b>336</b> 413-525 (2000) </a>.





          <li><a href="atompaw-4.0.0.13.tar.gz" target="_blank">
      atompaw-4.0.0.13.tar.gz </a>(5.4mb) 2/15/2016  Marc Torrent
      introduced improvements to XML output for abinit and new option
      for command line determination of version.   MT also prepared
      some notes <a href="./atompaw.README.MacOSX" target="_blank">
      in text format</a> or
      <a href="./atompaw.README.MacOSX.pdf"> in PDF format </a>
      to install atompaw under MacOS.  In addition, NH made some slight
      changes to the UPF output for Quantum Espresso towards compatibility
      with libxc.</li>


      <li><a href="atompaw-4.0.0.12.tar.gz" target="_blank">
      atompaw-4.0.0.12.tar.gz </a>(5.4mb) 2/03/2015  Slight update of
      <a href="./newperiodictable" target="_blank"> new version </a>  of
      atompaw code. In additon to previous <a href="./archive.html"
       target="_blank">
      updates</a>,  added PBESOL output for Quantum Espresso interface
      and added interface for SOCORRO.</li>

    <li><a href="atompaw-3.1.0.3.tar.gz" target="_blank">
       atompaw-3.1.0.3.tar.gz </a>(3.8mb) 
      Updated version of <i> atompaw </i> code 
      (01/03/2014 and 09/18/2013 -- Marc Torrent and Francois Jollet 
             introduced 
         improvements to the XML and abinit dataset generation routines;
       07/09/2013 -- Marc Torrent introduced small corrections;
       06/22/2013 -- Marc Torrent and Francois Jollet added a new option
        for outputting a file in XML format according to the specifications
        set up by the 
        <a href="https://wiki.fysik.dtu.dk/gpaw/setups/pawxml.html#pawxml">
        GPAW group </a>.  The output file format is controled by
        a menu at the end of the dataset:
        2 for standard abinit output, 3 for quantum-espresso (UPF) output,
        4 for XML output.)
       

      (Older changes:
      09/20/2012 -- Yann Pouillon updated the autotools for constructing
      the tar file; 07/16/2012 -- Geoffrey Pourois
      corrected GIPAW portion of pwscfinterface.f90; 06/26/12 -- 
     NAWH corrected bug
      in pwscfinterface.f90 
      on top of previous revisions on
      06/13/12 and  04/14/12 by Marc Torrent), and previous changes
      10/03/11 by Marc Torrent and Yann Pouillon updating interface
      for use with LibXc.  This version is still compatible with
      pwscf including recent addition by 
      D. Ceresoli for gipaw calculations.
      The
      3.0+
      versions have  several  features due largely to the magic of 
      Marc Torrent (CEA, France) and Yann Pouillon (ETSF, Spain) including
      compatability for use with LibXC. </li>
      <ul>
        <BLOCKQUOTE><li> The code package now complies with linux 
              installation standards. </li>
        <li> Using new options in the input file, datasets 
               for use with <a href="http://www.abinit.org" target="_blank"> 
               <i>abinit</i></a> (replacing the need
               to run the separate atompaw2abinit code) or 
              <a href="http://www.pwscf.org" target="_blank"> 
              <i>pwscf, quantum-espresso</i></a> can be generated.  
              (For developing the UPF file for use with 
               <a href="http://www.pwscf.org" target="_blank"><i>pwscf</i></a>,
               help from Lorenzo Paulatto and Paolo Giannozzi 
                is gratefully acknowledged.) </li>
        <li> The use of atompaw with 
          <a href="http://www.tddft.org/programs/octopus/wiki/index.php/Libxc" target="_blank">
          LibXC</a> library of exchange-correlation 
              functionals are now possible for generating
          datasets for <a href="http://www.abinit.org" target="_blank"> 
           <i>abinit</i></a>. </li>
        <li> Details are given in the 
        <a href="./atompaw-usersguide.pdf" target="_blank">  
          user's guide </a> written by Marc Torrent. </li>
        <li> Some details concerning choices of the shapes of compensation 
         charge
          densities have been clarified as explained in a recent 
          <a href="http://www.wfu.edu/~natalie/papers/PAWform/index.html" target="_blank"> 
            publication</a>. </li>
        <li>Simple <a href="./plot" target="_blank"> 
          gnuplot scripts </a> are available to help analyze 
         some of the outputs of the atompaw program. </li>
        </BLOCKQUOTE>
      </ul>
  
    <li><a href="atompaw-4.0.0.10.tar.gz" target="_blank">
      atompaw-4.0.0.10.tar.gz </a>(5.4mb) 8/22/2014: Slight update of
      <a href="./newperiodictable" target="_blank"> new version </a>  
      atompaw code.   Corrections  from 4.0.0.8 version include
      updated versions of xml output as specified on 
      <a href="https://wiki.fysik.dtu.dk/gpaw/setups/pawxml.html#pawxml">
      "standard" </a> webpage.  
      Note that <a href="http://www.abinit.org"> <i><b>ABINIT</b></i></a>
      now reads xml formatted atomic data files as well as the abinit
      "proprietary" formatted atomic data files.  
      <a href="http://quantum-espresso.org">
      <i><b> Quantum Espresso</b></i></a> is continuing to use its 
      "UPF" formatted atomic data files.  With the use of keywords, the atompaw
      code can output the atomic data in any or all of these formats.
       Marc Torrent introduced some computer compatibility
      updates to the atompaw in version 4.0.0.10.   In addition,
      core-core and core-valence Fock exchange terms are
      provided in anticipation of their use in "exact" exchange and
      hybrid calculations (with thanks to Jens J&oslash;rgen Mortensen).
      (<i> Note: full testing of exchange terms
      is not yet complete.</i>)





<li>
<a href="atompaw-4.0.0.8.tar.gz" target="_blank">
      atompaw-4.0.0.8.tar.gz </a>(5.4mb)
      <a href="./newperiodictable" target="_blank"> new version </a>
      atompaw code with solver and coretailpoints bugs corrected;
    updated version of xml interface (but not completely tested. (01/17/2014).

<li>
Atompaw versions with minor changes:

<a href="atompaw-3.1.0.0.tar.gz"> atompaw-3.1.0.0.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.9.tar.gz"> atompaw-3.0.1.9.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.8.tar.gz"> atompaw-3.0.1.8.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.7.tar.gz"> atompaw-3.0.1.7.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.6.tar.gz"> atompaw-3.0.1.6.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.5.tar.gz"> atompaw-3.0.1.5.tar.gz </a>(3.7mb)
<a href="atompaw-3.0.1.4.tar.gz"> atompaw-3.0.1.4.tar.gz </a>(3.7mb)
<li>
<a href="atompaw-3.0.1.3.tar.gz"> atompaw-3.0.1.3.tar.gz </a>(3.7mb)
Updated version of <i> atompaw </i> code (10/03/11 -- slightly
revised from 6/22/11 with addition to  pwscf interface introduced by
D. Ceresoli for gipaw calculations. (Thanks also to G. Lopez and T.
Thonhauser for help checking the resulting UPF files.)
The
3.0+
versions have  several  features due largely to the magic of
Marc Torrent (CEA, France) and Yann Pouillon (ETSF, Spain) including
compatability for use with LibXC.
<li>
<a href="atompaw-3.0.1.2.tar.gz"> atompaw-3.0.1.2.tar.gz </a>(3.7mb) 
Updated version of <i> atompaw </i> code (6/22/11 -- very slightly 
revised from 12/13/10 with small correction to pwscf interface (NAWH),
compability for use with LibXC (MT), and version number updates (YP)). The
3.0+
versions have  several  features due largely to the magic of 
Marc Torrent (CEA, France) and Yann Pouillon (ETSF, Spain).
<li> 
<a href="atompaw_2.3.tar.gz"> atompaw_2.3.tar.gz </a>(0.7mb) 
Updated version of <i> atompaw </i> code (05/12/10) with new options
for compatibility with <a href="http://www.abinit.org"> <i>abinit</i> </a> code 
versions 6.1 and higher.  The update concerns the treatment of the
argument of the exchange-correlation functional. The <i>pwpaw</i> code
and optionally the
6.1+ <i>abinit</i> code construct the argument from the
 sum of the pseudo core and valence densities. Older
versions and optionally 6.1+ <i>abinit</i> also include the valence compensation
charge in this argument which can occasionally cause numerical 
inaccuracies as discussed in the <a href="./PAWformman.pdf"> manuscript
draft</a>. 
An "unofficial" version of the <a href="atompaw2abinit.tgz">
Atompaw2abinit (3.3.0) </a>
 program written by Marc Torrent and Francois Jollet
is consistent with this new version
of <i>atompaw</i>. This program  prepares input to 
<i> abinit</i> from either [atom].atomicdata
or [atom].[exc].xml outputs of <i>atompaw</i></a>; the official version 
  is  available from the 
<a href='http://www.abinit.org'><i>abinit</i></a> website. 
Marc Torrent has prepared a 
<a href="http://www.abinit.org/documentation/helpfiles/for-v6.0/tutorial/lesson_
paw2/atompawUG.pdf/">
manual </a> for the <i>atompaw</i> code (with the new options).
General notes on  the <i>atompaw</i>, <i>pwpaw</i>, and <i>abinit</i>
 implementations of PAW  are
also 
<a href="./notes/atompaw/atompawEqns.pdf"> available</a>.  (For some general
comments see <a href="archive.html#notesarchive">archive</a>.)
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze 
some of the outputs of the atompaw program.
<li>
<a href="atompaw_2.2.patched.tgz"> atompaw_2.2.patched.tgz </a>(0.2mb)
Updated version of <i> atompaw </i> code (09/02/07) with
examples designated as Version 2.2.patched, as slightly modified by
Marc Torrent 1/09 to stabilize GGA functional. This version as a few additional
options developed by Marc Torrent with one more example.
The tar file contains an "unofficial" version of the Atompaw2abinit (3.2.0)
 program to prepare input to <i> abinit</i> from either [atom].atomicdata
or [atom].[exc].xml outputs of atompaw written by
 Marc Torrent and Francois Jollet which is also available from the
<a href='http://www.abinit.org'>abinit</a> website.
Marc Torrent has prepared a
<a href='http://www.abinit.org/PAW/AtomPAW2Abinit-Manual-html/atompawUG.pdf'>
manual </a> for the atompaw code (with the new options).
General notes on  the atompaw, pwpaw, and abinit implementations of PAW  are
also
<a href="./notes/atompaw/atompawEqns.pdf"> available</a>.  (For some general
comments see <a href="archive.html#notesarchive">archive</a>.)
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze
some of the outputs of the atompaw program.
<li> <a href="pwpaw_2.3.tar.gz"> pwpaw_2.3.tar.gz </a> (0.4 mb) Updated
12/23/2009 with very minor changes to some plotting routines and the
addition of a BSD license file.
<li>
<a href="122106pwpaw.tar.gz">
 122106pwpaw.tar.gz </a> (0.4mb)
This version of <i>pwpaw</i> is compatibible with the new options in
<i>atompaw</i>, reading the new form of the [atom].atomicdata files.  In
addition,
  several bugs were fixed.
<li>
<a href="atompaw_2.2.tar.gz"> atompaw_2.2.tar.gz </a>(0.2mb)
Updated version of <i> atompaw </i> code (09/02/07) with
examples designated as Version 2.2. This version as a few additional
options developed by Marc Torrent with one more example.
The tar file contains an "unofficial" version of the Atompaw2abinit (3.2.0)
 program to prepare input to <i> abinit</i> from either [atom].atomicdata
or [atom].[exc].xml outputs of atompaw written by
 Marc Torrent and Francois Jollet which is also available from the
<a href='http://www.abinit.org'>abinit</a> website.
Marc Torrent has prepared a
<a href='http://www.abinit.org/PAW/AtomPAW2Abinit-Manual-html/atompawUG.pdf'>
manual </a> for the atompaw code (with the new options).
General notes on  the atompaw, pwpaw, and abinit implementations of PAW  are
also
<a href="./notes/atompaw/atompawEqns.pdf"> available</a>.  (For some general
comments see <a href="archive.html#notesarchive">archive</a>.)
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze
some of the outputs of the atompaw program.
<li>
<a href="atompaw_2.1.tar.gz"> atompaw_2.1.tar.gz</a>(0.2mb)
Updated version of <i> atompaw </i> code (02/16/07) with
examples designated as Version 2.1 which has very minor corrections
from 2.0 version.
 In addition to several bug fixes
(including a significant error in the GGA equations),
Marc Torrent has introduced several new options.
<Li>
<a href="atompaw_2.0.tar.gz"> atompaw_2.0.tar.gz </a>(0.2mb)
Updated version of <i> atompaw </i> code with
examples designated as Version 2.0.  In addition to several bug fixes
(including a significant error in the GGA equations),
Marc Torrent has introduced several new options which are explained in a
manual available on the
 <a href='http://www.abinit.org'>abinit</a> web page.
The tar file contains an "unofficial" version of the Atompaw2abinit
 program to prepare input to <i> abinit</i> from either [atom].atomicdata
or [atom].[exc].xml outputs of atompaw.  A more complete version of
Atompaw2abinit  is
being developed by Marc Torrent and Francois Jollet will be available from the
<a href='http://www.abinit.org'>abinit</a> website.
In addition to the
<a href='http://www.abinit.org/PAW/AtomPAW2Abinit-Manual-html/atompawUG.pdf'>
manual </a> prepared by Marc Torrent, notes on the method are
<a href="./notes/atompaw/atompawEqns.pdf"> available</a>.  (For some general
comments see <a href="archive.html#notesarchive">archive</a>.)
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze
some of the outputs of the atompaw program.
<li>
<a href="071206atompaw.tar.gz">
071206atompaw.tar.gz</a> (1.8mb)
<a name="notesarchive">
Updated version of <i> atompaw </i> code with
examples.   Thanks to Marc Torrent and Francois Jollet, this version is
fully compatible with the  <a href='http://www.abinit.org'>abinit</a> code
and the tar file contains an "unofficial" version of the Atompaw2abinit
 program to prepare input to <i> abinit</i> from either [atom].atomicdata
or [atom].[exc].xml outputs of atompaw.  A more complete version of
Atompaw2abinit
being developed by Marc Torrent and will be available from the
<a href='http://www.abinit.org'>abinit</a> website.
The xml output from atompaw is now in compliance with the
<a href="http://dcwww.camp.dtu.dk/campos//pawxml/pawxml.xhtml">
pawxml.xhtml </a> standard maintained by Jens J&ouml;rgen Morgensen.
 Other changes include the capability
of using a logarithmic grid for evaluating and storing the radial functions,
the possibility of solving scalar relativistic equations instead of
the non-relativistic Schr&ouml;dinger equation, and the possibility
of choosing a squared sinc OR a Gaussian shape function for representing
the compensation charge (thanks to a suggestion by Peter Bl&ouml;chl).
[Note: Because of the way that the <i>abinit</i> code evaluates the
long-range Coulombic contributions, it is important that the
compensation charge density is well localized within the augmentation
sphere.  Therefore, in using the Gaussian shape function in the
<i>abinit</i> code, it is important to adjust the Gaussian width parameter
accordingly.]
In addition, we have simplified the keywords, while keeping the most common
old versions for backward compatibility.
As explained in the
<a href="./notes/atompaw/atompawEqns.pdf">
notes </a> listed below, the first step of the calculation is to determine
a local pseudopotential based on the norm-conserving  formalism of Troulliers
and Martins.   The second step is to construct basis and projector functions forthe smaller values of <i>l</i> according to either
<a href="http://www.pt.tu-clausthal.de/atp/">
Peter Bl&ouml;chl's </a> scheme
(keyword: BLOECHL)  or
<a href='http://www.physics.rutgers.edu/~dhv/uspp/'>
David Vanderbilt's </a> scheme (keyword: VANDERBILT) from the
ultra-soft pseudopotential formalism.
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze
some of the outputs of the atompaw program.</a>
<li ><a href="071206pwpaw.tar.gz">
 071206pwpaw.tar.gz </a> (2.8mb)
This version of <i>pwpaw</i> is compatibible with the new options in
<i>atompaw</i>, reading the new form of the [atom].atomicdata files.  In
addition,
  several bugs were fixed.

<li>
<a href="012406atompaw.tar.gz">
012406atompaw.tar.gz</a> (3.3mb)
Updated version of <i> atompaw </i> code with
examples.   In this version, we have added output to the [atom].atomicdata
file that can be used with <a href='http://www.abinit.org'>abinit</a> code.
A program to convert between <i>pwpaw</i> and <i> abinit</i> format is
being developed by Marc Torrent.   In the mean time, an temporary version
is <a href="unofficialatompaw2abinit.tar.gz"> available </a>.
In addition, we have added an option to calculate the basis and projector
functions in a way that is similar to the one used by David Vanderbilt's
<a href='http://www.physics.rutgers.edu/~dhv/uspp/'> ultra-soft
pseudopotentials</a>.   In our experience, this scheme can lead to
more rapidly converging calculations.   Details are given in the
<a href="./notes/atompaw/atompawEqns.pdf">
notes </a> listed below.   The xml output of this program needs work.
Simple <a href="./plot"> gnuplot scripts </a> are available to help analyze
some of the outputs of the atompaw program.
<li><a href="020406pwpaw.tgz">
 020406pwpaw.tgz </a> (3.1mb)  In this version, thanks to help from Alan
Wright, we have improved the G-space convergence by ensuring that the
pseudowavefunctions are represented by all wavevectors such that
|k+G| < G<sub>cut</sub>.   We now have excellent agreement with socorro
and abinit total  using the same projector and basis functions and the
same values of  G<sub>cut</sub>.   Minor improvements to example files.
<li><a href="012406pwpaw.tar.gz">
 012406pwpaw.tar.gz </a> (4.0mb)  Updated version
of <i>pwpaw</i> code with examples. Minor changed for compatibility with
the additional output in the [atom].atomicdata files.   Also reduced redundant
output for restart files.
<a href="111605atompaw.tar.gz">
111605atompaw.tar.gz</a> (1.8mb)
Slightly updated version of <i> atompaw </i> code with
examples.  (See <a href='./archive.html'> 01/08/05 version</a>
for details.)  Small bug fix since 082605 version.
<li><a href="082605pwpaw.tar.gz">
 082605pwpaw.tar.gz </a> (2.4 mb)  Revised version (08/26/05)
of <i>pwpaw</i> code with examples, implementing option to treat spin
dependence and also implementing more convenient restart files.
<li><a href="./notes/newatompawupdate.pdf"> Notes on new atompaw
implementation</a>
<li> <a href="082605atompaw.tar.gz">
082605atompaw.tar.gz</a> (1.8mb)
Slightly updated version of <i> atompaw </i> code with
examples.  (See <a href='./archive.html'> 01/08/05 version</a>
for details.)
<li> <a href="010805coretailatompaw.tar.gz">
010805coretailatompaw.tar.gz</a> (1mb)
New version (01/08/05) of <i>atompaw</i> code
 with examples.  New features include the use of
 a pseudo-core density to treat the frozen core density
 contributions which extend beyond rc as described
 in the updated <a href="./notes/coretail.pdf">notes</a>
 and the creation
 of output ([label].xml) in the xml format developed by the
 <a href="http://www.fsatom.org">Fsatom</a> group and described
 on the web page:
 <a href="http://www.fysik.dtu.dk/campos/atomic_setup/paw_setup.html">
 http://www.fysik.dtu.dk/campos/atomic_setup/paw_setup.html</a>.
 Also included is <i>xmlpaw</i> which uses Alan Tackett's fortran
 code to interpret the [label].xml file and create an input
 file [label].atomicdata.fromxml to be used by <i>pwpaw</i> program.
 The <i>xmlpaw</i> program is currently very primitive and
 limited to linear grid input.
<li><a href="010805coretailpwpaw.tar.gz">
 010805coretailpwpaw.tar.gz (0.3 mb) </a> Revised version (01/08/05)
of <i>pwpaw</i> code with example, implementing coretail contributions.
<li><a href="081304pwpaw.tar.gz"> 
 081304pwpaw.tar.gz (0.3 mb) </a> Revised version (08/13/04) 
of pwpaw code with minor bug fix.
<li> <a href="072204atompaw.tar.gz"> 
 072204atompaw.tar.gz (0.01 mb) </a> 
New version (07/22/04) of atompaw code
 with better 
options for constructing V<sub>loc</sub>(r)
and better coding, as described in these 
<a href="./notes/atompawupdate.pdf"> notes</a>.
  The resulting output
file [case].atomicdata contains a new entry with keyword 
'VLOCFUN' for listing the
radial form of V<sub>loc</sub>(r).  Tar file contains example
inputs and outputs for a few atoms.
<li><a href="071304pwpaw.tar.gz"> 
 071304pwpaw.tar.gz (0.3 mb) </a> Revised version (07/13/04) 
of pwpaw code 
for compatibility
with new atompaw output and a few bug fixes.
<li> <a href="071304atompaw.tar.gz"> 
 071304atompaw.tar.gz (0.01 mb) </a> 
<li> <a href="050704atompaw.tar.gz"> 050704atompaw.tar.gz (0.01 mb) </a>
<li><a href="043004pwpaw.tar.gz">  043004pwpaw.tar.gz (0.3 mb) </a>
<li> <a href="update070903.tar.gz"> update070903.tar.gz</a> (3.0 mb)
 Updated atompaw and pwpaw  codes (070903). 
Changes were introduced to atompaw by Francois Jollet and Marc Torrent 
for compatibility with 
the <a href="http://www.abinit.org"> Abinit code. </a> Also, the calculation 
of the Vloc amplitude was automated for the recommended procedure. Tar
file also contains some sample input and output data for atompaw 
and updated versions
of the gnuplot scripts. Correction to pwpaw code fixes bug in charges.f90
subroutine (thanks to HuangYuan). <br>
 <a href="absoftexecutables.tar.gz">
Executable files compiled with absoft Pro Fortran 8.0 </a> (2 mb).
<li> <a href="update011003.tar.gz"> update011003.tar.gz</a> (0.4 mb)
 Updated pwpaw and atompaw codes (011003).
Bug fix in Simpson's rule integrator; compatibility with absoft compiler. 
<li> <a href="update061302.tar.gz"> update061302.tar.gz</a> (0.4 mb)
Updated pwpaw code (06/13/02).
Bug fix in wavefunction storage routines; stabilization of
bandstructure mode.  atompaw code also included, but not changed since
051402 update.
<li> <a href="update051402.tar.gz"> update051402.tar.gz</a> (0.3 mb)
Updated atompaw and pwpaw codes (05/14/02).
 Minor corrections to GGA code; VHartree matrix elements symmetrized. 
Unused libraries removed.
<li> <a href="update010802.tar.gz"> update010802.tar.gz</a> (0.7 mb)
<font color="black">Updated atompaw and pwpaw codes (01/08/02)
</font>. Corrects some convergence bugs in atompaw and some gga bugs in pwpaw.
<li> <a href="gga.tar.gz"> gga.tar.gz</a>
 (1.9 mb)  Updated atompaw and pwpaw code with examples (8/25/01) 
 New version supports both LDA and GGA calculations.
<li> <a href="pwpawcodeupdate.050301.tar.gz"> pwpawcodeupdate.050301.tar.gz</a>
 (0.3 mb) <font color="black"> Updated source code (5/3/01) </font>
<li> <a href="atompaw.tar.gz">atompaw.tar.gz</a> (1.2 mb)
<li> <a href="pwpaw.tar.gz">pwpaw.tar.gz</a> (16.7 mb)
(Updated 11/10/00)
</ul>