Regarding nk* and nq*

[appalakondaiah]

Dear Admin,

I am running the bulk MX2 compounds and facing below error with epw.out. can you please suggest me the possible suggesition for further

Error in routine wigner_seitz (1):
weights do not add up to nk1*nk2*nk3
nkf1 = 6
nkf2 = 6
nkf3 = 4

nqf1 = 6
nqf2 = 6
nqf3 = 4

nk1 = 6
nk2 = 6
nk3 = 4

nq1 = 6
nq2 = 6
nq3 = 4
/
21 cartesian
0.000000000000000E+00 0.000000000000000E+00 0.000000000000000E+00 1.0000000
0.000000000000000E+00 0.000000000000000E+00 0.618359692346929E-01 1.0000000
0.000000000000000E+00 0.000000000000000E+00 -0.123671938469386E+00 1.0000000
0.944289008412737E-01 0.163555653804091E+00 0.000000000000000E+00 1.0000000
0.944289008412737E-01 0.163555653804091E+00 0.618359692346929E-01 1.0000000
0.944289008412737E-01 0.163555653804091E+00 -0.123671938469386E+00 1.0000000
0.188857801682547E+00 0.327111307608182E+00 0.000000000000000E+00 1.0000000
0.188857801682547E+00 0.327111307608182E+00 0.618359692346929E-01 1.0000000
0.188857801682547E+00 0.327111307608182E+00 -0.123671938469386E+00 1.0000000
-0.283286702523821E+00 -0.490666961412273E+00 0.000000000000000E+00 1.0000000
-0.283286702523821E+00 -0.490666961412273E+00 0.618359692346929E-01 1.0000000
-0.283286702523821E+00 -0.490666961412273E+00 -0.123671938469386E+00 1.0000000
0.283286702338488E+00 0.163555653804091E+00 0.000000000000000E+00 1.0000000
0.283286702338488E+00 0.163555653804091E+00 0.618359692346929E-01 1.0000000
0.283286702338488E+00 0.163555653804091E+00 -0.123671938469386E+00 1.0000000
0.377715603179762E+00 0.327111307608182E+00 0.000000000000000E+00 1.0000000
0.377715603179762E+00 0.327111307608182E+00 0.618359692346929E-01 1.0000000
0.377715603179762E+00 0.327111307608182E+00 -0.123671938469386E+00 1.0000000
0.566573404676977E+00 0.327111307608182E+00 0.000000000000000E+00 1.0000000
0.566573404676977E+00 0.327111307608182E+00 0.618359692346929E-01 1.0000000
0.566573404676977E+00 0.327111307608182E+00 -0.123671938469386E+00 1.0000000

[appalakondaiah]

Dear all,

As I continuing the previous post, I am requesting the possible solution for my error in epw calculations.

My Phonon calculations did with 24X24X8 Kmesh for SCF and 6X6X4 q-mesh for frequencies. I also follow same for epw-scf (24X24X8) and epw-nscf (216 Kpoints with 6X6X4 grid).

Error in routine wigner_seitz (1):
weights do not add up to nk1*nk2*nk3

Thanks in advance

with regards

[sponce]

Hello,

Can you show us your scf.in and nscf.in input file? Especially the k-point grid that you used.

Best,

Samuel

[appalakondaiah]

Dear Prof. Samuel,

Below I am giving the K-mesh details. Since, I did my phonons with 6X6X4 grid, I used the same grid in epw-nscf.in Kpoints by generating the utility of wannier90 (kmesh.pl)

epw-scf.in
K_POINTS automatic
24 24 8 0 0 0
epw-nscf.in
K_POINTS crystal
144
0.00000000 0.00000000 0.00000000 6.944444e-03
0.00000000 0.00000000 0.25000000 6.944444e-03
0.00000000 0.00000000 0.50000000 6.944444e-03
0.00000000 0.00000000 0.75000000 6.944444e-03
0.00000000 0.16666667 0.00000000 6.944444e-03
0.00000000 0.16666667 0.25000000 6.944444e-03
0.00000000 0.16666667 0.50000000 6.944444e-03
0.00000000 0.16666667 0.75000000 6.944444e-03
0.00000000 0.33333333 0.00000000 6.944444e-03
0.00000000 0.33333333 0.25000000 6.944444e-03
0.00000000 0.33333333 0.50000000 6.944444e-03
0.00000000 0.33333333 0.75000000 6.944444e-03
0.00000000 0.50000000 0.00000000 6.944444e-03
0.00000000 0.50000000 0.25000000 6.944444e-03
0.00000000 0.50000000 0.50000000 6.944444e-03
0.00000000 0.50000000 0.75000000 6.944444e-03
0.00000000 0.66666667 0.00000000 6.944444e-03
0.00000000 0.66666667 0.25000000 6.944444e-03
0.00000000 0.66666667 0.50000000 6.944444e-03
0.00000000 0.66666667 0.75000000 6.944444e-03
0.00000000 0.83333333 0.00000000 6.944444e-03
0.00000000 0.83333333 0.25000000 6.944444e-03
0.00000000 0.83333333 0.50000000 6.944444e-03
0.00000000 0.83333333 0.75000000 6.944444e-03
0.16666667 0.00000000 0.00000000 6.944444e-03
0.16666667 0.00000000 0.25000000 6.944444e-03
-----
----
--

with regards
S. Appalakondaiah

[sponce]

Hello,

I'm afraid I really don't have much suggestion. This is quite weird.
Maybe try with a different grid to see if the problem persists.

Best,

Samuel

[appalakondaiah]

Dear Prof. Samuel,


I changed the phonon grid and try to check epw calculations. Still I am facing the same error, which I am showing below. Kindly suggest me the possible way!!!

The .epb files have been correctly written


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine wigner_seitz (1):
weights do not add up to nk1*nk2*nk3
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

[appalakondaiah]

Dear all,

I am running the epw calculations for Bulk MX2 compounds and facing the below error for various sets of kpoint and qpoints grids.

===================================================================
irreducible q point # 14
===================================================================

Symmetries of small group of q: 12

Number of q in the star = 2
List of q in the star:
1 0.566573405 0.327111308 -0.123671938
2 -0.566573405 -0.327111308 0.123671938

q( 71 ) = ( 0.5665734 0.3271113 -0.1236719 )
q( 72 ) = ( -0.5665734 -0.3271113 0.1236719 )

Writing epmatq on .epb files


The .epb files have been correctly written


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine wigner_seitz (1):
weights do not add up to nk1*nk2*nk3
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Here with I am attaching all my inputs for phonon and epw calculations and kindly suggest me the possible way. Thanks in advance

Phonons-scf
&CONTROL
calculation = "scf",
pseudo_dir = "/home/konda/Work/PWSCF/Pseudo",
outdir = "/home/konda/Work/PWSCF/TMP/WS2/Phonons/Ground",
prefix='WS2'
tstress =.true.
tprnfor =.true.
etot_conv_thr = 1.0D-8
forc_conv_thr = 1.0D-6

/
&SYSTEM
ibrav = 0
A= 3.153200
nat = 6
ntyp = 2
ecutwfc =90
degauss=.005,
smearing='mp'
occupations='smearing',
input_dft='vdw-df2'
/
&electrons
conv_thr = 1.0d-10
/
ATOMIC_SPECIES
S 32.065 S.pz-bhs.UPF
W 183.84 W.pz-bhs.UPF
CELL_PARAMETERS (alat= 5.95868444)
0.882498183 -0.509510564 0.000000000
-0.000000000 1.019021127 0.000000000
0.000000000 0.000000000 4.042954337
ATOMIC_POSITIONS (crystal)
S 0.333333333 0.666666667 0.629080401
S 0.333333333 0.666666667 0.870919599
S 0.666666667 0.333333333 0.129080401
S 0.666666667 0.333333333 0.370919599
W 0.333333333 0.666666667 0.250000000
W 0.666666667 0.333333333 0.750000000
K_POINTS automatic
24 24 8 0 0 0

Phonons
Phonons of WS2
&inputph
tr2_ph=1.0d-15,
alpha_mix = 0.3,
prefix='$NAME',
outdir='$TMP',
ldisp=.true.,
fildvscf = 'dvscf'
fildyn='$NAME.dyn',
nq1=6, nq2=6, nq3=2
/

epw-scf.in
&CONTROL
calculation = "scf",
pseudo_dir = "/home/konda/Work/PWSCF/Pseudo",
outdir = "/home/konda/Work/PWSCF/TMP/WS2/Phonons/Ground/epw",
wf_collect = .true.
prefix='WS2'
tstress =.true.
tprnfor =.true.
etot_conv_thr = 1.0D-8
forc_conv_thr = 1.0D-6

/
&SYSTEM
ibrav = 0
A= 3.153200
nat = 6
ntyp = 2
ecutwfc =90
degauss=.005,
smearing='mp'
occupations='smearing',
/
&electrons
conv_thr = 1.0d-10
/
ATOMIC_SPECIES
S 32.065 S.pz-bhs.UPF
W 183.84 W.pz-bhs.UPF
CELL_PARAMETERS (alat= 5.95868444)
0.882498183 -0.509510564 0.000000000
-0.000000000 1.019021127 0.000000000
0.000000000 0.000000000 4.042954337
ATOMIC_POSITIONS (crystal)
S 0.333333333 0.666666667 0.629080401
S 0.333333333 0.666666667 0.870919599
S 0.666666667 0.333333333 0.129080401
S 0.666666667 0.333333333 0.370919599
W 0.333333333 0.666666667 0.250000000
W 0.666666667 0.333333333 0.750000000
K_POINTS automatic
24 24 8 0 0 0
epw-nscf
&CONTROL
calculation = "scf",
pseudo_dir = "/home/konda/Work/PWSCF/Pseudo",
outdir = "/home/konda/Work/PWSCF/TMP/WS2/Phonons/Ground/epw",
prefix='WS2'
tstress =.true.
tprnfor =.true.
wf_collect = .true.
etot_conv_thr = 1.0D-8
forc_conv_thr = 1.0D-6

/
&SYSTEM
ibrav = 0
A= 3.153200
nat = 6
ntyp = 2
ecutwfc =90
degauss=.005,
smearing='mp'
occupations='smearing',
/
&electrons
conv_thr = 1.0d-10
/
ATOMIC_SPECIES
S 32.065 S.pz-bhs.UPF
W 183.84 W.pz-bhs.UPF
CELL_PARAMETERS (alat= 5.95868444)
0.882498183 -0.509510564 0.000000000
-0.000000000 1.019021127 0.000000000
0.000000000 0.000000000 4.042954337
ATOMIC_POSITIONS (crystal)
S 0.333333333 0.666666667 0.629080401
S 0.333333333 0.666666667 0.870919599
S 0.666666667 0.333333333 0.129080401
S 0.666666667 0.333333333 0.370919599
W 0.333333333 0.666666667 0.250000000
W 0.666666667 0.333333333 0.750000000
K_POINTS crystal
72
0.00000000 0.00000000 0.00000000 1.388889e-02
0.00000000 0.00000000 0.50000000 1.388889e-02
0.00000000 0.16666667 0.00000000 1.388889e-02
0.00000000 0.16666667 0.50000000 1.388889e-02
0.00000000 0.33333333 0.00000000 1.388889e-02
0.00000000 0.33333333 0.50000000 1.388889e-02
0.00000000 0.50000000 0.00000000 1.388889e-02
0.00000000 0.50000000 0.50000000 1.388889e-02
0.00000000 0.66666667 0.00000000 1.388889e-02
0.00000000 0.66666667 0.50000000 1.388889e-02
0.00000000 0.83333333 0.00000000 1.388889e-02
0.00000000 0.83333333 0.50000000 1.388889e-02
0.16666667 0.00000000 0.00000000 1.388889e-02
0.16666667 0.00000000 0.50000000 1.388889e-02
0.16666667 0.16666667 0.00000000 1.388889e-02
------
-----
------

epw.in
--
&inputepw
prefix = 'WS2'
amass(1) = 32.05,
amass(2) = 183.84,
outdir = '/home/konda/Work/PWSCF/TMP/WS2/Phonons/Ground/epw/'

iverbosity = 0

elph = .true.
epbwrite = .true.
epbread = .false.

etf_mem = .false.

epwwrite = .true.
epwread = .false.

nbndsub = 22
wannierize = .true.
num_iter = 100
proj(1) = 'S:l=1'
proj(2) = 'W:l=2'
dis_win_min = -1.0
dis_froz_min = 8.0
dis_froz_max = 9.0
dis_win_max = 12.0
wdata(1) = 'bands_plot = .true.'
wdata(2) = 'begin kpoint_path'
wdata(3) = 'G 0.00 0.00 0.00 M 0.50 0.00 0.00'
wdata(4) = 'M 0.50 0.00 0.00 K 0.33 0.33 0.00'
wdata(5) = 'K 0.33 0.33 0.00 G 0.00 0.00 0.00'
wdata(6) = 'G 0.00 0.00 0.00 A 0.00 0.00 0.50'
wdata(7) = 'A 0.00 0.00 0.50 L 0.50 0.00 0.50'
wdata(8) = 'L 0.50 0.00 0.50 H 0.33 0.33 0.50'
wdata(9) = 'end kpoint_path'
wdata(10) = 'bands_plot_format = xmgr'
wdata(11) = 'guiding_centres = TRUE '

elecselfen = .true.
phonselfen = .true.
a2f = .false.

parallel_k = .true.
parallel_q = .false.


fsthick = 4.0 ! eV
eptemp = 300 ! K (same as PRB 76, 165108)
degaussw = 0.06802849 ! eV


dvscf_dir = '/home/konda/Work/PWSCF/TMP/WS2/Phonons/Ground/save'
filukk = './WS2.ukk'

nkf1 = 6
nkf2 = 6
nkf3 = 2

nqf1 = 6
nqf2 = 6
nqf3 = 2

nk1 = 6
nk2 = 6
nk3 = 2

nq1 = 6
nq2 = 6
nq3 = 2

/
14 cartesian
0.000000000 0.000000000 0.000000000 1.0000000
0.000000000 0.000000000 -0.123671938 1.0000000
0.094428901 0.163555654 0.000000000 1.0000000
0.094428901 0.163555654 -0.123671938 1.0000000
0.188857802 0.327111308 0.000000000 1.0000000
0.188857802 0.327111308 -0.123671938 1.0000000
-0.283286703 -0.490666961 0.000000000 1.0000000
-0.283286703 -0.490666961 -0.123671938 1.0000000
0.283286702 0.163555654 0.000000000 1.0000000
0.283286702 0.163555654 -0.123671938 1.0000000
0.377715603 0.327111308 0.000000000 1.0000000
0.377715603 0.327111308 -0.123671938 1.0000000
0.566573405 0.327111308 0.000000000 1.0000000
0.566573405 0.327111308 -0.123671938 1.0000000

[appalakondaiah]

Dear Admin/all

Can any one suggest the possible solution for my previous query!!! Thanks in advance

with regards
S. Appalakondaiah

[sponce]

Dear appalakondaiah,

You can have a look in the routine "appalakondaiah".

This routine was originally taken from the Wannier90 code.

This routines tries to find the Wigner size points. In the case of a very distorted cell (also 2D materials like your could be problematic),
the routines has troube finding all of them.

At the moment it looks in a supercells from -2,2 times the primitive cell. This means 5^3=125 times more. In practice it is from 0,4 which is equivalent.

You can increase this (will be slower) to for example 0,6 (equivalent to -3,3). You will therefore have 7^3 times more.

Note that "63" is the middle of 125 and is hardcoded so you will need to adapt it if you change it.

Hope this helps,

Best,
Samuel

[ywchoi]

I encountered the same problem when calculating graphene using EPW.
EPW dies at wigner_seitz.f90 saying

weights do not add up to nk1*nk2*nk3

But, this problem can be fixed by using less restrictive value of epsilon at the line 132 in wigner_seitz.f90

Code: Select all

DO while ( abs(dist(i)-mindist).le.eps7 .and. i.lt.125 )

, where eps7 = 1.0e-7_dp.

I found this fix by looking the source of wannier90.x; hamiltonian_wigner_seitz subroutine in hamiltonian.F90.
But I am not sure why this works.

Thank you.