[Pw_forum] U calculation error
Peng Chen
pchen at ion.chem.utk.edu
Sun Jul 29 22:01:08 CEST 2012
Dear Dr. Matteo,
Yes, I ran the LSDA+U example successfully, the only difference is that
the U of Ni converged to 6.0 not 5.0.
For my calculation, if I use n1=n2=n3=2, the result is
208 NaN 6.899999999999315E-003 2.066666666665403E-003.
If n1=n2=n3=3, the size of Umat.out will continue to increase > 2G. And I
have to stop it before the space is used up.
The potential and wfcs are saved and I used ecutwfc=100 to make the scf
energy converged to 0.001Ry. I can decrease it if that will help to get
correct U.
On Sun, Jul 29, 2012 at 1:17 PM, Matteo Cococcioni <matteo at umn.edu> wrote:
> Dear Peng,
>
> I advise you to run the example available from the QE webpage and try
> to reproduce the same results.
>
> On Sun, Jul 29, 2012 at 12:00 PM, Peng Chen <pchen at ion.chem.utk.edu>
> wrote:
> > Dear Dr. Matteo,
> >
> > I compiled the code with the parameters in make.sys. The n1=n2=n3=1 case
> is
> > done without error, but the results for Ni and O are negative which
> should
> > not be correct.
>
>
> ok then maybe you can try to increase them one per time and see at
> what level the code stops with error?
>
>
> > 26 -14.8678837915617 7.56505581785392 -0.681000339374195
> > I plot dn.1.da.1.dat and dn0.1.da.1.dat, the data in dn0.1.da.1.dat are
> not
> > linear.
>
>
> this is not related to the error message of r.x but is not right. if
> you want reliable results you have to get these plots linear.
>
>
>
> The scf calculation is converged to 0.001 Ry. So I am not sure what
> > can cause the problem.
>
>
> yes probably this is not enough. the inputs look fine (I haven't
> looked inside the response matrix) but you have to make sure the scf
> potential and wfc are saved after the unperturbed scf run and every
> perturbed calculation starts from them.
> do you really need 100 Ry cut off with US PP?
>
> Matteo
>
>
> >
> > I've attached the Umat_afm.1.out which may contain some useful
> information
> > for you. I also put below the input for scf and perturbation on Ni
> > (alpha=-0.15,-0.07,0.0,0.07,0.15). Same calculations are done for V and
> O as
> > well with perturbed V or O atom translated to the origin of the unit
> cell.
> >
> >
> > scf calculation input:
> >
> > &control
> > calculation='scf'
> > restart_mode='from_scratch',
> > prefix='nvoprim.afm.ucalc.Ni',
> > pseudo_dir = '/home/pchen/espresso-5.0/pseudo/',
> > outdir='/data/pchen/qe/'
> > wf_collect=.TRUE.
> > /
> > &system
> > ibrav=0, celldm(1)=11.19096,
> > nat=26, ntyp=6,
> > nspin = 2, starting_magnetization(1)=0.6,
> > starting_magnetization(2)=-0.6, starting_magnetization(3)=0.5,
> > starting_magnetization(4)=-0.5,
> > ecutwfc = 100, ecutrho = 800,
> > occupations ='smearing',
> > smearing ='mv',
> > degauss = 0.005,nbnd=135
> > lda_plus_u = .true.
> > Hubbard_U(1)= 1.d-20
> > Hubbard_U(2)= 1.d-20
> > Hubbard_U(3)= 1.d-20
> > Hubbard_U(4)= 1.d-20
> > Hubbard_U(5)= 1.d-20
> > Hubbard_U(6)= 1.d-20
> > /
> > &electrons
> > conv_thr = 1.0e-7
> > mixing_beta = 0.3
> > /
> > CELL_PARAMETERS
> > 0.5 0.96015 0
> > 0.5 -0.96015 0
> > 0 0 1.38889
> > ATOMIC_SPECIES
> > Ni1 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni2 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni3 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni4 58.6934 Ni.pz-n-rrkjus.UPF
> > V 50.9415 V.pz-spn-rrkjus.UPF
> > O 15.9994 O.pz-n-rrkjus.UPF
> > ATOMIC_POSITIONS crystal
> > Ni1 0.000000000 0.000000000 0.000000000
> > Ni2 0.500000000 0.500000000 0.500000000
> > Ni3 0.122863290 0.378190177 0.749934361
> > Ni3 0.877136710 0.621809823 0.250065639
> > Ni4 0.377137388 0.121809201 0.249932658
> > Ni4 0.622862612 0.878190799 0.750067342
> > V 0.373618754 0.623612138 0.115867980
> > V 0.126381501 0.876388862 0.615868582
> > V 0.626381246 0.376387862 0.884132020
> > V 0.873618499 0.123611138 0.384131418
> > O 0.241839036 0.756304999 0.223613779
> > O 0.258160617 0.743693910 0.723611944
> > O 0.758160964 0.243695001 0.776386221
> > O 0.741839383 0.256306090 0.276388056
> > O -0.000998332 1.000227439 0.243205274
> > O 0.500997862 0.499772508 0.743208489
> > O 1.000998332 -0.000227439 0.756794726
> > O 0.499002138 0.500227492 0.256791511
> > O 0.381099639 0.150692971 0.006796703
> > O 0.118898786 0.349304325 0.506798051
> > O 0.149581603 0.378555571 0.991989919
> > O 0.350419534 0.121445697 0.491991443
> > O 0.618900361 0.849307029 0.993205297
> > O 0.881101214 0.650695675 0.493203949
> > O 0.850418397 0.621444429 0.008012081
> > O 0.649580466 0.878554303 0.508010557
> >
> > K_POINTS (automatic)
> > 5 5 5 1 1 1
> >
> > Perturbation on Ni with alpha=0.15 input:
> >
> > &control
> > calculation='scf'
> > restart_mode='from_scratch',
> > prefix='nvoprim.afm.ucalc.Ni',
> > pseudo_dir = '/home/pchen/espresso-5.0/pseudo/',
> > outdir='/data/pchen/qe/'
> > wf_collect=.TRUE.
> > /
> > &system
> > ibrav=0, celldm(1)=11.19096,
> > nat=26, ntyp=6,
> > nspin = 2, starting_magnetization(1)=0.6,
> > starting_magnetization(2)=-0.6, starting_magnetization(3)=0.5,
> > starting_magnetization(4)=-0.5,
> > ecutwfc = 100, ecutrho = 800,
> > occupations ='smearing',
> > smearing ='mv',
> > degauss = 0.005,nbnd=135
> > lda_plus_u = .true.
> > Hubbard_U(1)= 1.d-20
> > Hubbard_U(2)= 1.d-20
> > Hubbard_U(3)= 1.d-20
> > Hubbard_U(4)= 1.d-20
> > Hubbard_U(5)= 1.d-20
> > Hubbard_U(6)= 1.d-20
> >
> > Hubbard_alpha(1)= 0.15
> > /
> > &electrons
> > conv_thr = 1.0e-7
> > mixing_beta = 0.3
> > startingpot='file'
> > startingwfc='file'
> > diago_thr_init=5.09E-11,
> > /
> > CELL_PARAMETERS
> > 0.5 0.96015 0
> > 0.5 -0.96015 0
> > 0 0 1.38889
> > ATOMIC_SPECIES
> > Ni1 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni2 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni3 58.6934 Ni.pz-n-rrkjus.UPF
> > Ni4 58.6934 Ni.pz-n-rrkjus.UPF
> > V 50.9415 V.pz-spn-rrkjus.UPF
> > O 15.9994 O.pz-n-rrkjus.UPF
> > ATOMIC_POSITIONS crystal
> > Ni1 0.000000000 0.000000000 0.000000000
> > Ni2 0.500000000 0.500000000 0.500000000
> > Ni3 0.122863290 0.378190177 0.749934361
> > Ni3 0.877136710 0.621809823 0.250065639
> > Ni4 0.377137388 0.121809201 0.249932658
> > Ni4 0.622862612 0.878190799 0.750067342
> > V 0.373618754 0.623612138 0.115867980
> > V 0.126381501 0.876388862 0.615868582
> > V 0.626381246 0.376387862 0.884132020
> > V 0.873618499 0.123611138 0.384131418
> > O 0.241839036 0.756304999 0.223613779
> > O 0.258160617 0.743693910 0.723611944
> > O 0.758160964 0.243695001 0.776386221
> > O 0.741839383 0.256306090 0.276388056
> > O -0.000998332 1.000227439 0.243205274
> > O 0.500997862 0.499772508 0.743208489
> > O 1.000998332 -0.000227439 0.756794726
> > O 0.499002138 0.500227492 0.256791511
> > O 0.381099639 0.150692971 0.006796703
> > O 0.118898786 0.349304325 0.506798051
> > O 0.149581603 0.378555571 0.991989919
> > O 0.350419534 0.121445697 0.491991443
> > O 0.618900361 0.849307029 0.993205297
> > O 0.881101214 0.650695675 0.493203949
> > O 0.850418397 0.621444429 0.008012081
> > O 0.649580466 0.878554303 0.508010557
> >
> > K_POINTS (automatic)
> > 5 5 5 1 1 1
> >
> >
> >
> > On Sun, Jul 29, 2012 at 11:27 AM, Matteo Cococcioni <matteo at umn.edu>
> wrote:
> >>
> >> Dear Peng,
> >>
> >> I'm not very expert of compiling codes. You have to make sure that all
> >> the libraries are linked correctly for the code to work. I assume you
> >> have QE installed on the same computer. You could give a look in the
> >> make.sys to check what is the correct way to link them. If everything
> >> is consistent, then you have to find out where the code is failing.
> >> have you tried to run with n1 = n2 = n3 = 1? what did you get?
> >>
> >> Matteo
> >>
> >>
> >>
> >> On Sat, Jul 28, 2012 at 9:10 PM, Peng Chen <pchen at ion.chem.utk.edu>
> wrote:
> >> > Dear Dr. Matteo,
> >> >
> >> > I used command listed below to compile resp_mat.f90. No error in
> >> > compiling.
> >> > I am not sure if I did right in choosing the libs.
> >> > ifort -o r.x resp_mat.f90 -L"/data/apps/intel/11.1.072/mkl/lib/em64t"
> >> > -lmkl_lapack95_lp64 -lmkl_intel_lp64 -lmkl_mc3 -lmkl_sequential
> >> > -lmkl_core
> >> >
> >> > The libs in mkl are shown below:
> >> > libmkl_avx.so libmkl_intel_sp2dp.so
> >> > libmkl_blacs_ilp64.a libmkl_intel_thread.a
> >> > libmkl_blacs_intelmpi_ilp64.a libmkl_intel_thread.so
> >> > libmkl_blacs_intelmpi_ilp64.so libmkl_lapack95_ilp64.a
> >> > libmkl_blacs_intelmpi_lp64.a libmkl_lapack95_lp64.a
> >> > libmkl_blacs_intelmpi_lp64.so libmkl_lapack.so
> >> > libmkl_blacs_lp64.a libmkl_mc3.so
> >> > libmkl_blacs_openmpi_ilp64.a libmkl_mc.so
> >> > libmkl_blacs_openmpi_lp64.a libmkl_p4n.so
> >> > libmkl_blacs_sgimpt_ilp64.a libmkl_pgi_thread.a
> >> > libmkl_blacs_sgimpt_lp64.a libmkl_pgi_thread.so
> >> > libmkl_blas95_ilp64.a libmkl_scalapack_ilp64.a
> >> > libmkl_blas95_lp64.a libmkl_scalapack_ilp64.so
> >> > libmkl_cdft_core.a libmkl_scalapack_lp64.a
> >> > libmkl_core.a libmkl_scalapack_lp64.so
> >> > libmkl_core.so libmkl_sequential.a
> >> > libmkl_def.so libmkl_sequential.so
> >> > libmkl_gf_ilp64.a libmkl_solver_ilp64.a
> >> > libmkl_gf_ilp64.so libmkl_solver_ilp64_sequential.a
> >> > libmkl_gf_lp64.a libmkl_solver_lp64.a
> >> > libmkl_gf_lp64.so libmkl_solver_lp64_sequential.a
> >> > libmkl_gnu_thread.a libmkl_vml_avx.so
> >> > libmkl_gnu_thread.so libmkl_vml_def.so
> >> > libmkl_intel_ilp64.a libmkl_vml_mc2.so
> >> > libmkl_intel_ilp64.so libmkl_vml_mc3.so
> >> > libmkl_intel_lp64.a libmkl_vml_mc.so
> >> > libmkl_intel_lp64.so libmkl_vml_p4n.so
> >> > libmkl_intel_sp2dp.a locale
> >> >
> >> >
> >> > Some errors in output:
> >> > forrtl: severe (174): SIGSEGV, segmentation fault occurred
> >> > Image PC Routine Line
> >> > Source
> >> > r.x 0000000000445659 Unknown Unknown
> >> > Unknown
> >> > r.x 000000000040A363 Unknown Unknown
> >> > Unknown
> >> > r.x 000000000040338C Unknown Unknown
> >> > Unknown
> >> > libc.so.6 00002B85B9495994 Unknown Unknown
> >> > Unknown
> >> > r.x 0000000000403289 Unknown Unknown
> >> > Unknown
> >> >
> >> >
> >> > On Sat, Jul 28, 2012 at 10:38 AM, Matteo Cococcioni <matteo at umn.edu>
> >> > wrote:
> >> >>
> >> >> Dear Peng,
> >> >>
> >> >> nothing looks wrong in your input. have you tried with n1, n2, n3 all
> >> >> equal to 1? are you sure you compiled correctly?
> >> >>
> >> >> Matteo
> >> >>
> >> >>
> >> >>
> >> >> On Fri, Jul 27, 2012 at 11:46 PM, Peng Chen <pchen at ion.chem.utk.edu>
> >> >> wrote:
> >> >> > Dear Dr. Matteo,
> >> >> >
> >> >> > The primitive cell has 26 atoms with 3 different types. The pos and
> >> >> > resp_mat.in files are shown below:
> >> >> >
> >> >> > 5.6087 10.79 0.0
> >> >> > 5.6087 -10.79 0.0
> >> >> > 0.0 0.0 15.570873
> >> >> > 0 0 0 1
> >> >> > 0.5 0.5 0.5 -1
> >> >> > 0.12286329 0.378190177 0.749934361 1
> >> >> > 0.87713671 0.621809823 0.250065639 1
> >> >> > 0.377137388 0.121809201 0.249932658 -1
> >> >> > 0.622862612 0.878190799 0.750067342 -1
> >> >> > 0.373618754 0.623612138 0.11586798 0
> >> >> > 0.126381501 0.876388862 0.615868582 0
> >> >> > 0.626381246 0.376387862 0.88413202 0
> >> >> > 0.873618499 0.123611138 0.384131418 0
> >> >> > 0.241839036 0.756304999 0.223613779 0
> >> >> > 0.258160617 0.74369391 0.723611944 0
> >> >> > 0.758160964 0.243695001 0.776386221 0
> >> >> > 0.741839383 0.25630609 0.276388056 0
> >> >> > -0.000998332 1.000227439 0.243205274 0
> >> >> > 0.500997862 0.499772508 0.743208489 0
> >> >> > 1.000998332 -0.000227439 0.756794726 0
> >> >> > 0.499002138 0.500227492 0.256791511 0
> >> >> > 0.381099639 0.150692971 0.006796703 0
> >> >> > 0.118898786 0.349304325 0.506798051 0
> >> >> > 0.149581603 0.378555571 0.991989919 0
> >> >> > 0.350419534 0.121445697 0.491991443 0
> >> >> > 0.618900361 0.849307029 0.993205297 0
> >> >> > 0.881101214 0.650695675 0.493203949 0
> >> >> > 0.850418397 0.621444429 0.008012081 0
> >> >> > 0.649580466 0.878554303 0.508010557 0
> >> >> >
> >> >> > resp_mat.in :
> >> >> > &input_mat
> >> >> > ntyp = 3
> >> >> > na(1) = 6
> >> >> > na(2) = 4
> >> >> > na(3) = 16
> >> >> > nalfa = 5
> >> >> > magn=.true.
> >> >> > filepos = 'pos'
> >> >> > back = 'no'
> >> >> > filednda = 'file_afm'
> >> >> > n1 = 2
> >> >> > n2 = 2
> >> >> > n3 = 2
> >> >> > &end
> >> >> >
> >> >> >
> >> >> >
> >> >> > On Fri, Jul 27, 2012 at 3:02 PM, Matteo Cococcioni <matteo at umn.edu
> >
> >> >> > wrote:
> >> >> >>
> >> >> >> Dear Peng,
> >> >> >>
> >> >> >> can you provide info about your system? maybe it is larger than
> the
> >> >> >> maximum size allowed by r.x?
> >> >> >>
> >> >> >> Matteo
> >> >> >>
> >> >> >> On Fri, Jul 27, 2012 at 12:41 PM, Peng Chen <
> pchen at ion.chem.utk.edu>
> >> >> >> wrote:
> >> >> >> > Dear All,
> >> >> >> >
> >> >> >> > There are some errors when I run r.x to calculate the U. I used
> >> >> >> > Espresso-5.0
> >> >> >> > and follow the steps in LSDA+U example. I am not sure what could
> >> >> >> > be
> >> >> >> > wrong.
> >> >> >> > And if you need other input/output, please let me know.
> >> >> >> >
> >> >> >> > Image PC Routine Line
> Source
> >> >> >> > r.x 0822DACC Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0822C265 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 082071FA Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DDFDA Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081FA33B Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081F9C5C Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804E38F Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804824A Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0823805D Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 08048131 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > forrtl: severe (8): internal consistency check failure, file
> >> >> >> > ./src/libfor/for_wseq_lis.c, line 635
> >> >> >> > Image PC Routine Line
> Source
> >> >> >> > r.x 0822DACC Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0822C265 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 082071FA Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DD664 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081F9AF7 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804E38F Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804824A Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0823805D Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 08048131 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > forrtl: severe (174): SIGSEGV, segmentation fault occurred
> >> >> >> > Image PC Routine Line
> Source
> >> >> >> > r.x 0804DAED Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804824A Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0823805D Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 08048131 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > forrtl: severe (174): SIGSEGV, segmentation fault occurred
> >> >> >> > Image PC Routine Line
> Source
> >> >> >> > r.x 08253657 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081F4087 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DCDB7 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DF354 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DD672 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 081DF5E8 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > Unknown FFFFE600 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0804824A Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 0823805D Unknown Unknown
> >> >> >> > Unknown
> >> >> >> > r.x 08048131 Unknown Unknown
> >> >> >> > Unknown
> >> >> >> >
> >> >> >> >
> >> >> >> > --
> >> >> >> > Best Regards.
> >> >> >> > Peng
> >> >> >> >
> >> >> >> > _______________________________________________
> >> >> >> > Pw_forum mailing list
> >> >> >> > Pw_forum at pwscf.org
> >> >> >> > http://www.democritos.it/mailman/listinfo/pw_forum
> >> >> >> >
> >> >> >>
> >> >> >>
> >> >> >>
> >> >> >> --
> >> >> >> Matteo Cococcioni
> >> >> >> Department of Chemical Engineering and Materials Science,
> >> >> >> University of Minnesota
> >> >> >> 421 Washington Av. SE
> >> >> >> Minneapolis, MN 55455
> >> >> >> Tel. +1 612 624 9056 Fax +1 612 626 7246
> >> >> >> _______________________________________________
> >> >> >> Pw_forum mailing list
> >> >> >> Pw_forum at pwscf.org
> >> >> >> http://www.democritos.it/mailman/listinfo/pw_forum
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> > --
> >> >> > Best Regards.
> >> >> > Peng
> >> >> >
> >> >> > _______________________________________________
> >> >> > Pw_forum mailing list
> >> >> > Pw_forum at pwscf.org
> >> >> > http://www.democritos.it/mailman/listinfo/pw_forum
> >> >> >
> >> >>
> >> >>
> >> >>
> >> >> --
> >> >> Matteo Cococcioni
> >> >> Department of Chemical Engineering and Materials Science,
> >> >> University of Minnesota
> >> >> 421 Washington Av. SE
> >> >> Minneapolis, MN 55455
> >> >> Tel. +1 612 624 9056 Fax +1 612 626 7246
> >> >> _______________________________________________
> >> >> Pw_forum mailing list
> >> >> Pw_forum at pwscf.org
> >> >> http://www.democritos.it/mailman/listinfo/pw_forum
> >> >
> >> >
> >> >
> >> >
> >> > --
> >> > Best Regards.
> >> > Peng
> >> >
> >> > _______________________________________________
> >> > Pw_forum mailing list
> >> > Pw_forum at pwscf.org
> >> > http://www.democritos.it/mailman/listinfo/pw_forum
> >> >
> >>
> >>
> >>
> >> --
> >> Matteo Cococcioni
> >> Department of Chemical Engineering and Materials Science,
> >> University of Minnesota
> >> 421 Washington Av. SE
> >> Minneapolis, MN 55455
> >> Tel. +1 612 624 9056 Fax +1 612 626 7246
> >> _______________________________________________
> >> Pw_forum mailing list
> >> Pw_forum at pwscf.org
> >> http://www.democritos.it/mailman/listinfo/pw_forum
> >
> >
> >
> >
> > --
> > Best Regards.
> > Peng
> >
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> >
>
>
>
> --
> Matteo Cococcioni
> Department of Chemical Engineering and Materials Science,
> University of Minnesota
> 421 Washington Av. SE
> Minneapolis, MN 55455
> Tel. +1 612 624 9056 Fax +1 612 626 7246
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
--
Best Regards.
Peng
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20120729/5ad91c94/attachment.html>
More information about the users
mailing list