Dear EPW admin and users,
Recently I used EPW to calculate the EPC of electrons in graphene, and met with the problem in wannierization if EPW with large spread. Although EPW used wannier90 through library mode, it is still very strange that EPW and wannier90.x give different results even they have the same configurations.
1)EPW
---------------
epw.in
----------------
graphene
&inputepw
prefix = 'graphene'
amass(1) = 12.0107
outdir = './'
iverbosity = 0
elph = .true.
epbwrite = .true.
epbread = .false.
elecselfen = .true.
phonselfen = .true.
a2f = .false.
epwwrite = .true.
epwread = .false.
etf_mem = .true.
nbndsub = 8
nbndskip = 0
wannierize = .true.
num_iter = 4000
dis_win_max = 20
dis_froz_max= 10
dis_win_min = -30
dis_froz_min= -10
iprint = 1
proj(1) = 'C:sp2;pz'
wdata(1) = 'dis_mix_ratio = 0.2'
wdata(2) = 'guiding_centres=.true.'
wdata(3) = 'num_print_cycles=10'
wdata(4) = 'dis_num_iter= 4000'
parallel_k = .true.
parallel_q = .false.
fsthick = 20.0 ! eV
eptemp = 300 ! K (same as PRB 76, 165108)
degaussw = 0.01 ! eV
dvscf_dir = '../phonons/save'
filukk = './graphene.ukk'
filqf = 'meshes/path2.dat'
nkf1 = 20
nkf2 = 20
nkf3 = 1
nk1 = 10
nk2 = 10
nk3 = 1
nqf1 = 60
nqf2 = 60
nqf3 = 1
nq1 = 10
nq2 = 10
nq3 = 1
/
14 cartesian
0.000000000 0.000000000 0.000000000 1
0.000000000 0.115470054 0.000000000 1
0.000000000 0.230940108 0.000000000 1
0.000000000 0.346410162 0.000000000 1
0.000000000 0.461880215 0.000000000 1
0.000000000 -0.577350269 0.000000000 1
0.100000000 0.173205081 0.000000000 1
0.100000000 0.288675135 0.000000000 1
0.100000000 0.404145188 0.000000000 1
0.100000000 0.519615242 0.000000000 1
0.200000000 0.346410162 0.000000000 1
0.200000000 0.461880215 0.000000000 1
0.200000000 0.577350269 0.000000000 1
0.300000000 0.519615242 0.000000000 1
------------------------------------------------------------------------
and the graphene.wout produced by EPW
--------------------------------------------------------------
Wannier90 is running in LIBRARY MODE
Setting up k-point neighbours...
------
SYSTEM
------
Lattice Vectors (Ang)
a_1 2.435274 0.000000 0.000000
a_2 -1.217637 2.109009 0.000000
a_3 0.000000 0.000000 9.741094
Unit Cell Volume: 50.03039 (Ang^3)
Reciprocal-Space Vectors (Ang^-1)
b_1 2.580074 1.489606 0.000000
b_2 0.000000 2.979213 0.000000
b_3 0.000000 0.000000 0.645018
*----------------------------------------------------------------------------*
| Site Fractional Coordinate Cartesian Coordinate (Ang) |
+----------------------------------------------------------------------------+
| C 1 0.33333 0.66667 0.00000 | -0.00001 1.40601 0.00000 |
| C 2 0.66667 0.33333 0.00000 | 1.21765 0.70300 0.00000 |
*----------------------------------------------------------------------------*
------------
K-POINT GRID
------------
Grid size = 10 x 10 x 1 Total points = 100
*---------------------------------- MAIN ------------------------------------*
| Number of Wannier Functions : 8 |
| Number of input Bloch states : 8 |
| Output verbosity (1=low, 5=high) : 1 |
| Timing Level (1=low, 5=high) : 1 |
| Optimisation (0=memory, 3=speed) : 3 |
| Length Unit : Ang |
| Post-processing setup (write *.nnkp) : T |
| Using Gamma-only branch of algorithms : F |
*----------------------------------------------------------------------------*
*------------------------------- WANNIERISE ---------------------------------*
| Total number of iterations : 4000 |
| Number of CG steps before reset : 5 |
| Trial step length for line search : 2.000 |
| Convergence tolerence : 0.100E-09 |
| Convergence window : -1 |
| Iterations between writing output : 10 |
| Iterations between backing up to disk : 100 |
| Write r^2_nm to file : F |
| Write xyz WF centres to file : F |
| Write on-site energies <0n|H|0n> to file : F |
| Use guiding centre to control phases : T |
| Use phases for initial projections : F |
| Iterations before starting guiding centres: 0 |
| Iterations between using guiding centres : 1 |
*----------------------------------------------------------------------------*
Time to read parameters 0.001 (sec)
*---------------------------------- K-MESH ----------------------------------*
+----------------------------------------------------------------------------+
| Distance to Nearest-Neighbour Shells |
| ------------------------------------ |
| Shell Distance (Ang^-1) Multiplicity |
| ----- ----------------- ------------ |
| 1 0.297921 6 |
| 2 0.516015 6 |
| 3 0.595843 6 |
| 4 0.645018 2 |
| 5 0.710497 12 |
| 6 0.788226 12 |
| 7 0.826027 12 |
| 8 0.878110 12 |
| 9 0.893764 6 |
| 10 1.018503 24 |
| 11 1.032030 6 |
| 12 1.074170 12 |
+----------------------------------------------------------------------------+
| The b-vectors are chosen automatically |
| SVD found small singular value, Rejecting this shell and trying the next |
| The following shells are used: 1, 4 |
+----------------------------------------------------------------------------+
| Shell # Nearest-Neighbours |
| ----- -------------------- |
| 1 6 |
| 4 2 |
+----------------------------------------------------------------------------+
| Completeness relation is fully satisfied [Eq. (B1), PRB 56, 12847 (1997)] |
+----------------------------------------------------------------------------+
| b_k Vectors (Ang^-1) and Weights (Ang^2) |
| ---------------------------------------- |
| No. b_k(x) b_k(y) b_k(z) w_b |
| --- -------------------------------- -------- |
| 1 0.000000 0.297921 0.000000 3.755569 |
| 2 0.258007 0.148961 0.000000 3.755569 |
| 3 0.000000 -0.297921 0.000000 3.755569 |
| 4 0.258007 -0.148961 0.000000 3.755569 |
| 5 -0.258007 -0.148961 0.000000 3.755569 |
| 6 -0.258007 0.148961 0.000000 3.755569 |
| 7 0.000000 0.000000 0.645018 1.201782 |
| 8 0.000000 0.000000 -0.645018 1.201782 |
+----------------------------------------------------------------------------+
| b_k Directions (Ang^-1) |
| ----------------------- |
| No. x y z |
| --- -------------------------------- |
| 1 0.000000 0.297921 0.000000 |
| 2 0.258007 0.148961 0.000000 |
| 3 0.258007 -0.148961 0.000000 |
| 4 0.000000 0.000000 0.645018 |
+----------------------------------------------------------------------------+
Time to write kmesh 0.035 (sec)
graphene.nnkp written.
Time to write kmesh 0.035 (sec)
Finished setting up k-point neighbours.
Exiting wannier_setup in wannier90 20:31:15
Resuming Wannier90 at 20:31:21
------
SYSTEM
------
Lattice Vectors (Ang)
a_1 2.435274 0.000000 0.000000
a_2 -1.217637 2.109009 0.000000
a_3 0.000000 0.000000 9.741094
Unit Cell Volume: 50.03039 (Ang^3)
Reciprocal-Space Vectors (Ang^-1)
b_1 2.580074 1.489606 0.000000
b_2 0.000000 2.979213 0.000000
b_3 0.000000 0.000000 0.645018
*----------------------------------------------------------------------------*
| Site Fractional Coordinate Cartesian Coordinate (Ang) |
+----------------------------------------------------------------------------+
| C 1 0.33333 0.66667 0.00000 | -0.00001 1.40601 0.00000 |
| C 2 0.66667 0.33333 0.00000 | 1.21765 0.70300 0.00000 |
*----------------------------------------------------------------------------*
------------
K-POINT GRID
------------
Grid size = 10 x 10 x 1 Total points = 100
*---------------------------------- MAIN ------------------------------------*
| Number of Wannier Functions : 8 |
| Number of input Bloch states : 8 |
| Output verbosity (1=low, 5=high) : 1 |
| Timing Level (1=low, 5=high) : 1 |
| Optimisation (0=memory, 3=speed) : 3 |
| Length Unit : Ang |
| Post-processing setup (write *.nnkp) : T |
| Using Gamma-only branch of algorithms : F |
*----------------------------------------------------------------------------*
*------------------------------- WANNIERISE ---------------------------------*
| Total number of iterations : 4000 |
| Number of CG steps before reset : 5 |
| Trial step length for line search : 2.000 |
| Convergence tolerence : 0.100E-09 |
| Convergence window : -1 |
| Iterations between writing output : 10 |
| Iterations between backing up to disk : 100 |
| Write r^2_nm to file : F |
| Write xyz WF centres to file : F |
| Write on-site energies <0n|H|0n> to file : F |
| Use guiding centre to control phases : T |
| Use phases for initial projections : F |
| Iterations before starting guiding centres: 0 |
| Iterations between using guiding centres : 1 |
*----------------------------------------------------------------------------*
Time to read parameters 0.001 (sec)
*---------------------------------- K-MESH ----------------------------------*
+----------------------------------------------------------------------------+
| Distance to Nearest-Neighbour Shells |
| ------------------------------------ |
| Shell Distance (Ang^-1) Multiplicity |
| ----- ----------------- ------------ |
| 1 0.297921 6 |
| 2 0.516015 6 |
| 3 0.595843 6 |
| 4 0.645018 2 |
| 5 0.710497 12 |
| 6 0.788226 12 |
| 7 0.826027 12 |
| 8 0.878110 12 |
| 9 0.893764 6 |
| 10 1.018503 24 |
| 11 1.032030 6 |
| 12 1.074170 12 |
+----------------------------------------------------------------------------+
| The b-vectors are chosen automatically |
| SVD found small singular value, Rejecting this shell and trying the next |
| The following shells are used: 1, 4 |
+----------------------------------------------------------------------------+
| Shell # Nearest-Neighbours |
| ----- -------------------- |
| 1 6 |
| 4 2 |
+----------------------------------------------------------------------------+
| Completeness relation is fully satisfied [Eq. (B1), PRB 56, 12847 (1997)] |
+----------------------------------------------------------------------------+
| b_k Vectors (Ang^-1) and Weights (Ang^2) |
| ---------------------------------------- |
| No. b_k(x) b_k(y) b_k(z) w_b |
| --- -------------------------------- -------- |
| 1 0.000000 0.297921 0.000000 3.755569 |
| 2 0.258007 0.148961 0.000000 3.755569 |
| 3 0.000000 -0.297921 0.000000 3.755569 |
| 4 0.258007 -0.148961 0.000000 3.755569 |
| 5 -0.258007 -0.148961 0.000000 3.755569 |
| 6 -0.258007 0.148961 0.000000 3.755569 |
| 7 0.000000 0.000000 0.645018 1.201782 |
| 8 0.000000 0.000000 -0.645018 1.201782 |
+----------------------------------------------------------------------------+
| b_k Directions (Ang^-1) |
| ----------------------- |
| No. x y z |
| --- -------------------------------- |
| 1 0.000000 0.297921 0.000000 |
| 2 0.258007 0.148961 0.000000 |
| 3 0.258007 -0.148961 0.000000 |
| 4 0.000000 0.000000 0.645018 |
+----------------------------------------------------------------------------+
Time to get kmesh 0.031 (sec)
Time to project overlaps 0.006 (sec)
*------------------------------- WANNIERISE ---------------------------------*
+--------------------------------------------------------------------+<-- CONV
| Iter Delta Spread RMS Gradient Spread (Ang^2) Time |<-- CONV
+--------------------------------------------------------------------+<-- CONV
------------------------------------------------------------------------------
Initial State
WF centre and spread 1 ( -0.137538, 1.322977, 0.125092 ) 79.24337563
WF centre and spread 2 ( -0.175059, 1.349527, -0.199757 ) 82.85307173
WF centre and spread 3 ( -0.349815, 0.965515, 0.161264 ) 80.01026740
WF centre and spread 4 ( -0.116468, 1.434882, -0.166926 ) 94.10592761
WF centre and spread 5 ( 1.724800, 0.837941, -0.225388 ) 84.57915002
WF centre and spread 6 ( 0.569208, 0.656591, 0.169822 ) 80.39051522
WF centre and spread 7 ( 1.944299, 0.803295, -0.015008 ) 78.61691338
WF centre and spread 8 ( 1.225933, 0.619000, -0.238771 ) 101.54432285
Sum of centres and spreads ( 4.685359, 7.989729, -0.389671 ) 681.34354385
0 0.681E+03 0.0000000000 681.3435438466 4.40 <-- CONV
O_D= 481.8726052 O_OD= 0.1256001 O_TOT= 681.3435438 <-- SPRD
.......................
Final State
WF centre and spread 1 ( -1.577396, 2.644543, -0.139091 ) 25.50495412
WF centre and spread 2 ( -0.497653, 0.424520, -0.119889 ) 25.46362802
WF centre and spread 3 ( -0.025438, 3.269723, 0.079791 ) 25.53219332
WF centre and spread 4 ( 2.381811, 1.065457, -0.014307 ) 26.25036349
WF centre and spread 5 ( 0.209067, 1.145020, -0.129766 ) 25.46432705
WF centre and spread 6 ( 0.069150, 0.701874, -0.083083 ) 25.50396858
WF centre and spread 7 ( 1.308099, 0.838071, -0.008784 ) 25.16043567
WF centre and spread 8 ( -1.992943, 4.316313, -0.051324 ) 25.62866500
Sum of centres and spreads ( -0.125303, 14.405521, -0.466453 ) 204.50853525
Spreads (Ang^2) Omega I = 199.345338555
================ Omega D = 5.042648147
Omega OD = 0.120548544
Final Spread (Ang^2) Omega Total = 204.508535246
2)wannier90.x
----------------------------------------
graphene.win
----------------------------------------
num_wann = 8
dis_win_max = 19.2d0
dis_froz_max = 9.8d0
dis_win_min = -30.0
dis_froz_min = -10
dis_num_iter = 4000
num_iter = 4000
num_print_cycles = 10
guiding_centres = true
----------------------------------------
results by wannier90.x
----------------------------------------
******************************************************************************
* -> Using CODATA 2006 constant values *
* (http://physics.nist.gov/cuu/Constants/index.html) *
* -> Using Bohr value from CODATA *
******************************************************************************
------
SYSTEM
------
Lattice Vectors (Ang)
a_1 2.435274 0.000000 0.000000
a_2 -1.217637 2.109009 0.000000
a_3 0.000000 0.000000 9.741094
Unit Cell Volume: 50.03039 (Ang^3)
Reciprocal-Space Vectors (Ang^-1)
b_1 2.580074 1.489606 0.000000
b_2 0.000000 2.979213 0.000000
b_3 0.000000 0.000000 0.645018
*----------------------------------------------------------------------------*
| Site Fractional Coordinate Cartesian Coordinate (Ang) |
+----------------------------------------------------------------------------+
| C 1 0.33333 0.66667 0.00000 | -0.00000 1.40601 0.00000 |
| C 2 0.66667 0.33333 0.00000 | 1.21764 0.70300 0.00000 |
*----------------------------------------------------------------------------*
------------
K-POINT GRID
------------
Grid size = 10 x 10 x 1 Total points = 100
*---------------------------------- MAIN ------------------------------------*
| Number of Wannier Functions : 8 |
| Number of input Bloch states : 8 |
| Output verbosity (1=low, 5=high) : 1 |
| Timing Level (1=low, 5=high) : 1 |
| Optimisation (0=memory, 3=speed) : 3 |
| Length Unit : Ang |
| Post-processing setup (write *.nnkp) : F |
| Using Gamma-only branch of algorithms : F |
*----------------------------------------------------------------------------*
*------------------------------- WANNIERISE ---------------------------------*
| Total number of iterations : 4000 |
| Number of CG steps before reset : 5 |
| Trial step length for line search : 2.000 |
| Convergence tolerence : 0.100E-09 |
| Convergence window : -1 |
| Iterations between writing output : 10 |
| Iterations between backing up to disk : 100 |
| Write r^2_nm to file : F |
| Write xyz WF centres to file : F |
| Write on-site energies <0n|H|0n> to file : F |
| Use guiding centre to control phases : T |
| Use phases for initial projections : F |
| Iterations before starting guiding centres: 0 |
| Iterations between using guiding centres : 1 |
*----------------------------------------------------------------------------*
*-------------------------------- PLOTTING ----------------------------------*
| Plotting interpolated bandstructure : T |
| Number of K-path sections : 3 |
| Divisions along first K-path section : 100 |
| Output format : gnuplot |
| Output mode : s-k |
*----------------------------------------------------------------------------*
| K-space path sections: |
| From: G 0.000 0.000 0.000 To: M 0.500 0.000 0.000 |
| From: M 0.500 0.000 0.000 To: K 0.333 0.333 0.000 |
| From: K 0.333 0.333 0.000 To: G 0.000 0.000 0.000 |
*----------------------------------------------------------------------------*
Time to read parameters 0.006 (sec)
*---------------------------------- K-MESH ----------------------------------*
+----------------------------------------------------------------------------+
| Distance to Nearest-Neighbour Shells |
| ------------------------------------ |
| Shell Distance (Ang^-1) Multiplicity |
| ----- ----------------- ------------ |
| 1 0.297921 6 |
| 2 0.516015 6 |
| 3 0.595843 6 |
| 4 0.645018 2 |
| 5 0.710497 12 |
| 6 0.788226 12 |
| 7 0.826027 12 |
| 8 0.878110 12 |
| 9 0.893764 6 |
| 10 1.018503 24 |
| 11 1.032030 6 |
| 12 1.074170 12 |
+----------------------------------------------------------------------------+
| The b-vectors are chosen automatically |
| SVD found small singular value, Rejecting this shell and trying the next |
| The following shells are used: 1, 4 |
+----------------------------------------------------------------------------+
| Shell # Nearest-Neighbours |
| ----- -------------------- |
| 1 6 |
| 4 2 |
+----------------------------------------------------------------------------+
| Completeness relation is fully satisfied [Eq. (B1), PRB 56, 12847 (1997)] |
+----------------------------------------------------------------------------+
| b_k Vectors (Ang^-1) and Weights (Ang^2) |
| ---------------------------------------- |
| No. b_k(x) b_k(y) b_k(z) w_b |
| --- -------------------------------- -------- |
| 1 0.000000 0.297921 0.000000 3.755569 |
| 2 0.258007 0.148961 0.000000 3.755569 |
| 3 0.000000 -0.297921 0.000000 3.755569 |
| 4 0.258007 -0.148961 0.000000 3.755569 |
| 5 -0.258007 -0.148961 0.000000 3.755569 |
| 6 -0.258007 0.148961 0.000000 3.755569 |
| 7 0.000000 0.000000 0.645018 1.201782 |
| 8 0.000000 0.000000 -0.645018 1.201782 |
+----------------------------------------------------------------------------+
| b_k Directions (Ang^-1) |
| ----------------------- |
| No. x y z |
| --- -------------------------------- |
| 1 0.000000 0.297921 0.000000 |
| 2 0.258007 0.148961 0.000000 |
| 3 0.258007 -0.148961 0.000000 |
| 4 0.000000 0.000000 0.645018 |
+----------------------------------------------------------------------------+
*============================================================================*
| MEMORY ESTIMATE |
| Maximum RAM allocated during each phase of the calculation |
*============================================================================*
| Wannierise: 2.25 Mb |
| plot_wannier: 2.25 Mb |
*----------------------------------------------------------------------------*
Starting a new Wannier90 calculation ...
Time to get kmesh 0.034 (sec)
Reading overlaps from graphene.mmn : Created on 15Jun2017 at 16:49:51
Reading projections from graphene.amn : Created on 15Jun2017 at 16:49:45
Time to read overlaps 0.051 (sec)
Writing checkpoint file graphene.chk... done
*------------------------------- WANNIERISE ---------------------------------*
+--------------------------------------------------------------------+<-- CONV
| Iter Delta Spread RMS Gradient Spread (Ang^2) Time |<-- CONV
+--------------------------------------------------------------------+<-- CONV
------------------------------------------------------------------------------
Initial State
WF centre and spread 1 ( -0.433795, 1.803607, -0.014418 ) 8.70308783
WF centre and spread 2 ( -0.167329, 0.864178, -0.009530 ) 8.82930575
WF centre and spread 3 ( 0.564482, 1.471991, 0.013204 ) 8.81027830
WF centre and spread 4 ( -0.000003, 1.406008, -0.000000 ) 2.60475809
WF centre and spread 5 ( 1.223409, 1.285185, 0.060689 ) 16.31688587
WF centre and spread 6 ( 0.549627, 0.392434, -0.017376 ) 16.78684016
WF centre and spread 7 ( 1.670977, 0.328055, -0.127535 ) 16.68099359
WF centre and spread 8 ( 1.217645, 0.702998, -0.000000 ) 5.79229066
Sum of centres and spreads ( 4.625013, 8.254454, -0.094966 ) 84.52444025
0 0.845E+02 0.0000000000 84.5244402487 0.09 <-- CONV
O_D= 14.4179397 O_OD= 52.0564116 O_TOT= 84.5244402 <-- SPRD
..............
Final State
WF centre and spread 1 ( -0.556796, 1.740959, -0.046432 ) 0.92305722
WF centre and spread 2 ( -0.017488, 0.693423, -0.074896 ) 1.06723144
WF centre and spread 3 ( 0.621087, 1.728085, 0.057316 ) 0.82930093
WF centre and spread 4 ( -0.014989, 1.391337, 0.076655 ) 1.91313954
WF centre and spread 5 ( 0.926325, 2.244232, -0.209130 ) 7.52932969
WF centre and spread 6 ( 0.325961, -0.047539, 0.132009 ) 7.90902604
WF centre and spread 7 ( 2.523059, 0.015260, 0.108331 ) 7.33729542
WF centre and spread 8 ( 1.220851, 0.699665, -0.034203 ) 5.59857190
Sum of centres and spreads ( 5.028010, 8.465422, 0.009649 ) 33.10695217
Spreads (Ang^2) Omega I = 18.050088887
================ Omega D = 0.965993790
Omega OD = 14.090869495
Final Spread (Ang^2) Omega Total = 33.106952172
wannier90 in EPW
Moderator: stiwari
Re: wannier90 in EPW
I had similar issue before. I still don't know how it works, but I solved the problem by using -npool in the nscf calculation for epw, and then enforce the same -npool in epw calculation. I am not sure if it will work for you, but maybe worth a try...
Re: wannier90 in EPW
jiachen wrote:I had similar issue before. I still don't know how it works, but I solved the problem by using -npool in the nscf calculation for epw, and then enforce the same -npool in epw calculation. I am not sure if it will work for you, but maybe worth a try...
It really works! Thank you very much!
Re: wannier90 in EPW
It is indeed a requirement. Glad it works for you now.
Best,
Samuel
Best,
Samuel
Prof. Samuel Poncé
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com
Chercheur qualifié F.R.S.-FNRS / Professeur UCLouvain
Institute of Condensed Matter and Nanosciences
UCLouvain, Belgium
Web: https://www.samuelponce.com