<?xml version="1.0" encoding="ISO-8859-1"?>
<?xml-stylesheet type="text/xsl" href="input_xx.xsl"?>
<!-- FILE AUTOMATICALLY CREATED: DO NOT EDIT, CHANGES WILL BE LOST -->
    
<input_description distribution="Quantum Espresso" package="PWscf" program="projwfc.x" >
   <toc>
   </toc>
   <intro>
<b>Purpose of projwfc.x:</b>
    projects wavefunctions onto orthogonalized atomic wavefunctions,
    calculates Lowdin charges, spilling parameter, projected DOS
    (separated into up and down components for lSDA)
    alternatively, computes the local DOS(E), integrated in volumes
    given in input

<b>Structure of the input data:</b>
============================

   <b>&amp;PROJWFC</b>
     ...
   <b>/</b>
   </intro>
   <namelist name="PROJWFC" >
      <var name="prefix" type="CHARACTER" >
         <info>
prefix of input file produced by <b>pw.x</b> (wavefunctions are needed)
         </info>
         <default> &apos;pwscf&apos;
         </default>
      </var>
      <var name="outdir" type="CHARACTER" >
         <info>
directory containing the input data, i.e. the same as in <b>pw.x</b>
         </info>
         <default>
value of the ESPRESSO_TMPDIR environment variable if set;
current directory (&apos;./&apos;) otherwise
         </default>
      </var>
      <var name="ngauss" type="INTEGER" >
         <default> 0
         </default>
         <info>
Type of gaussian broadening:
    0 ... Simple Gaussian (default)
    1 ... Methfessel-Paxton of order 1
   -1 ... &quot;cold smearing&quot; (Marzari-Vanderbilt-DeVita-Payne)
  -99 ... Fermi-Dirac function
         </info>
      </var>
      <var name="degauss" type="REAL" >
         <default> 0.0
         </default>
         <info> gaussian broadening, Ry (not eV!)
         </info>
      </var>
      <vargroup type="REAL" >
         <var name="Emin" >
         </var>
         <var name="Emax" >
         </var>
         <info> min &amp; max energy (eV) for DOS plot
         </info>
         <default> (band extrema)
         </default>
      </vargroup>
      <var name="DeltaE" type="REAL" >
         <info> energy grid step (eV)
         </info>
      </var>
      <var name="lsym" type="LOGICAL" >
         <default> .true.
         </default>
         <info>
if <b>.true.</b>  the projections are symmetrized,
           the partial density of states are computed
if <b>.false.</b> the projections are not symmetrized, the partial
           DOS can be computed only in the k-resolved case
         </info>
      </var>
      <var name="pawproj" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.</b> use PAW projectors and all-electron PAW basis
functions to calculate weight factors for the partial
densities of states. Following Bloechl, <a href="https://journals.aps.org/prb/abstract/10.1103/PhysRevB.50.17953">PRB 50, 17953 (1994)</a>,
Eq. (4 &amp; 6), the weight factors thus approximate the real
charge within the augmentation sphere of each atom.
Only for PAW, not implemented in the noncolinear case.
         </info>
      </var>
      <var name="filpdos" type="CHARACTER" >
         <info> prefix for output files containing PDOS(E)
         </info>
         <default> (value of <ref>prefix</ref> variable)
         </default>
      </var>
      <var name="filproj" type="CHARACTER" >
         <default> (standard output)
         </default>
         <info>
file containing the projections
         </info>
      </var>
      <var name="lwrite_overlaps" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.,</b> the overlap matrix of the atomic orbitals
prior to orthogonalization is written to the atomic_proj
datafile. Does not work together with linear-algebra
diagonalization: run as &quot;mpirun -np N projwfc.x -nd 1 ... &quot;
         </info>
      </var>
      <var name="lbinary_data" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.,</b> the atomic_proj datafile is written in binary fmt.
Currently disabled.
         </info>
      </var>
      <var name="kresolveddos" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.</b> the k-resolved DOS is computed: not summed over
all k-points but written as a function of the k-point index.
In this case all k-point weights are set to unity
         </info>
      </var>
      <var name="tdosinboxes" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.</b> compute the local DOS integrated in volumes

Volumes are defined as boxes with edges parallel to the unit cell,
containing the points of the (charge density) FFT grid included within
<ref>irmin</ref> and <ref>irmax</ref>, in the three dimensions:

from <ref>irmin</ref>(j,n) to <ref>irmax</ref>(j,n) for j=1,2,3 (n=1,<ref>n_proj_boxes</ref>).
         </info>
      </var>
      <var name="n_proj_boxes" type="INTEGER" >
         <default> 1
         </default>
         <info>
number of boxes where the local DOS is computed
         </info>
      </var>
      <var name="irmin(3,n_proj_boxes)" type="INTEGER" >
         <default> 1 for each box
         </default>
         <info>
first point of the given box

BEWARE: <ref>irmin</ref> is a 2D array of the form: <ref>irmin</ref>(3,<ref>n_proj_boxes</ref>)
         </info>
      </var>
      <var name="irmax(3,n_proj_boxes)" type="INTEGER" >
         <default> 0 for each box
         </default>
         <info>
last point of the given box;
( 0 stands for the last point in the FFT grid )

BEWARE: <ref>irmax</ref> is a 2D array of the form: <ref>irmax</ref>(3,<ref>n_proj_boxes</ref>)
         </info>
      </var>
      <var name="plotboxes" type="LOGICAL" >
         <default> .false.
         </default>
         <info>
if <b>.true.,</b> the boxes are written in output as <b>xsf</b> files with
3D datagrids, valued 1.0 inside the box volume and 0 outside
(visualize them as isosurfaces with isovalue 0.5)
         </info>
      </var>
   </namelist>
   <section title="Notes" >
      <subsection title="Format of output files" >
         <text>
Projections are written to standard output, and also to file
<ref>filproj</ref> if given as input.

The total DOS and the sum of projected DOS are written to file
&quot;filpdos&quot;.pdos_tot.

* The format for the collinear, spin-unpolarized case and the
  non-collinear, spin-orbit case is:
      E DOS(E) PDOS(E)
      ...

* The format for the collinear, spin-polarized case is:
      E DOSup(E) DOSdw(E)  PDOSup(E) PDOSdw(E)
      ...

* The format for the non-collinear, non spin-orbit case is:
      E DOS(E) PDOSup(E) PDOSdw(E)
      ...

In the collinear case and the non-collinear, non spin-orbit case
projected DOS are written to file &quot;filpdos&quot;.pdos_atm#N(X)_wfc#M(l),
where N = atom number , X = atom symbol, M = wfc number, l=s,p,d,f
(one file per atomic wavefunction found in the pseudopotential file)

* The format for the collinear, spin-unpolarized case is:
      E LDOS(E) PDOS_1(E) ... PDOS_2l+1(E)
      ...
  where LDOS = \sum m=1,2l+1 PDOS_m(E)
  and PDOS_m(E) = projected DOS on atomic wfc with component m

* The format for the collinear, spin-polarized case and the
  non-collinear, non spin-orbit case is as above with
  two components for both  LDOS(E) and PDOS_m(E)

In the non-collinear, spin-orbit case (i.e. if there is at least one
fully relativistic pseudopotential) wavefunctions are projected
onto eigenstates of the total angular-momentum.
Projected DOS are written to file &quot;filpdos&quot;.pdos_atm#N(X)_wfc#M(l_j),
where N = atom number , X = atom symbol, M = wfc number, l=s,p,d,f
and j is the value of the total angular momentum.
In this case the format is:
    E LDOS(E) PDOS_1(E) ... PDOS_2j+1(E)
    ...

If <ref>kresolveddos</ref>=.true., the k-point index is prepended
to the formats above, e.g. (collinear, spin-unpolarized case)
    ik E DOS(E) PDOS(E)

All DOS(E) are in states/eV plotted vs E in eV
         </text>
      </subsection>
      <subsection title="Orbital Order" >
         <text>
Order of m-components for each l in the output:

    1, cos(phi), sin(phi), cos(2*phi), sin(2*phi), .., cos(l*phi), sin(l*phi)

where phi is the polar angle:x=r cos(theta)cos(phi), y=r cos(theta)sin(phi)
This is determined in file Modules/ylmr2.f90 that calculates spherical harmonics.

for l=1:
  1 pz     (m=0)
  2 px     (real combination of m=+/-1 with cosine)
  3 py     (real combination of m=+/-1 with sine)

for l=2:
  1 dz2    (m=0)
  2 dzx    (real combination of m=+/-1 with cosine)
  3 dzy    (real combination of m=+/-1 with sine)
  4 dx2-y2 (real combination of m=+/-2 with cosine)
  5 dxy    (real combination of m=+/-2 with sine)
         </text>
      </subsection>
      <subsection title="Defining boxes for the Local DOS(E)" >
         <text>
Boxes are specified using the variables <ref>irmin</ref> and <ref>irmax</ref>:

FFT grid points are included from irmin(j,n) to irmax(j,n)
for j=1,2,3 and n=1,...,<ref>n_proj_boxes</ref>

<ref>irmin</ref> and <ref>irmax</ref> range from 1 to nr1 or nr2 or nr3

Values larger than nr1/2/3 or smaller than 1 are folded
to the unit cell.

If <ref>irmax</ref>&lt;<ref>irmin</ref> FFT grid points are included from 1 to irmax
and from irmin to nr1/2/3.
         </text>
      </subsection>
      <subsection title="Important notices" >
         <text>
The tetrahedron method is used if

    - the input data file has been produced by pw.x using the option
      occupations=&apos;tetrahedra&apos;, AND

    - a value for degauss is not given as input to namelist &amp;projwfc

* Gaussian broadening is used in all other cases:

    - if <ref>degauss</ref> is set to some value in namelist &amp;PROJWFC, that value
      (and the optional value for ngauss) is used

    - if <ref>degauss</ref> is NOT set to any value in namelist &amp;PROJWFC, the
      value of <ref>degauss</ref> and of <ref>ngauss</ref> are read from the input data
      file (they will be the same used in the pw.x calculations)

    - if <ref>degauss</ref> is NOT set to any value in namelist &amp;PROJWFC, AND
      there is no value of <ref>degauss</ref> and of <ref>ngauss</ref> in the input data
      file, <ref>degauss</ref>=<ref>DeltaE</ref> (in Ry) and <ref>ngauss</ref>=0 will be used


Obsolete variables, ignored:
    io_choice
    smoothing
         </text>
      </subsection>
   </section>
</input_description>