Re: [QE-users] SCAN functional in QE-6.5

2020-05-28 Thread Sergey Lisenkov 
+ ferrariruffino.fz@ Fabrizio, It seems I made a mistake but cannot figure out where. Because I cannot reproduce it. Anyway, you are correct - both 'input_dft = "scan"' and 'input_dft = "MGGA_X_SCAN + MGGA_C_SCAN" result in the same indexes:from PW output:   Exchange-correlation= MGGA_X_SCAN + MGGA_C_SCAN                           (   0   0   0   0   0 263 267) However, when one takes SCAN pseudopotentials from the library on QE website, and don't use 'input_dft = 'scan' ' option, then PW should take XC from pseudopotentials. But index is different: Exchange-correlation=  SLA  PW   SCAN SCAN                           (   1   4   0   0   0   5   0) Header of pseudopotential file is:==Generated using Vanderbilt code, version   7  3  2                              Author: Yi Yao     Generation date:    0    0    0                              Automatically converted from original format                                        0        The Pseudo was generated with a Non-Relativistic Calculationnl pn  l   occ               Rcut            Rcut US             E pseu0S  0  0  2.00      0.000      0.000      0.P  0  1  2.00      0.000      0.000      0.000    0                   Version Number   C                   Element   NC                  Norm - Conserving pseudopotential    F                  Nonlinear Core CorrectionSLA  PW   SCAN SCAN    SCAN Exchange-Correlation functional    4.000      Z valence    0.000      Total energy  0.000  0.000 Suggested cutoff for wfc and rho    0                  Max angular momentum component 1389                  Number of points in mesh    2    1             Number of Wavefunctions, Number of Projectors Wavefunctions         nl  l   occ                       0S  0  2.00                       0P  1  2.00=and this run fails with error: Error in routine tau_xc (5):     This case is not implemented I guess this is 'different' SCAN. thanks for your help, Sergey  ___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Fabrizio Ferrari 
The 'SCAN' name in qe is a compact notation that is directly translated
into the libxc indexes 263 and 267, equivalent to the functionals MGGA_X_SCAN
and MGGA_C_SCAN.
Regardless the name notation, if you find a line like this: (   0   0   0
0   0 263 267) in the output, after 'Exchange-correlation', it should be
ok, because it means that the two libxc functionals have been detected. If
not, let me know.
However it is clear that the best thing should be to understand why the
first notation does not work in your case.
Cheers

On Tue, May 26, 2020 at 11:24 AM Lorenzo Paulatto  wrote:

> > Your suggestion (input_dft = 'MGGA_X_SCAN MGGA_C_SCAN') worked.
>
>
> "Got the code to spit out a number" is not the same as "produced a valid
> result". I would wait for an answer from someone knowledgeable on this
> subject before spending too much CPU  time in this case!
>
> Or even better open an issue in the gitlab tracker
>  explaining why do you think that
> two equivalent cases are producing different outcomes. This was you'll
> have an answer for sure
>
> cheers
>
> > However, I still don't understand why others could use  (input_dft =
> > 'scan') without any problem - I searched archive and saw several
> > examples. Also, I used SCAN pseudopotentials from QE website, so in
> > principle I shouldn't declare input_dft option. And not declaring would
> > result in error (as I reported).
> > Thanks,
> > Sergey Lisenkov
> > 26.05.2020, 10:36, "Fabrizio Ferrari" :
> >
> > Hello Sergey,
> > that is strange. I have tried some simple tests too, but I could not
> > reproduce your issue. An equivalent way to activate the scan mgga
> > functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'
> > You can try with that, but if it still doesn't work, I think I would
> > need more details about your input in order to find the issue.
> > Cheers
> > On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov  > > wrote:
> >
> > Hello Fabrizio,
> > yes, it is compiled with libxc support. Without it, pw.x
> > complains immediately, that it should be compiled with Libxc
> > library (or something like that).
> > 25.05.2020, 17:21, "Fabrizio Ferrari"
> > mailto:ferrariruffino...@gmail.com
> >>:
> >
> > Hello,
> > in your case (scan) that error should come out only if libxc
> > has not been properly linked during compilation. Have you
> > tried runs with other libxc functionals (not necessarily
> > mGGA)? Do they work?
> > Just in case you want to check, to properly compile with
> > libxc, after the qe configuration, you have to add the
> > requested flags in the make.inc file at 'DFLAGS', 'IFLAGS'
> > and 'LD_LIBS' (you can follow the nearby comments).
> > Fabrizio
> > On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov
> > mailto:proff...@yandex.ru>> wrote:
> >
> > Hello,
> > I noticed that some people run SCAN calculations with QE
> > v.6.5. I tried this, but my test has failed.
> > I compiled 6.5 version with LibXC-4.3.4. When I activate
> > "input_dft = 'scan'" in my input file,
> > I get the following error just before first scf step:
> > 
> >   total cpu time spent up to now is1.0 secs
> >   Self-consistent Calculation
> >   iteration #  1 ecut=   100.00 Ry beta= 0.70
> >   Davidson diagonalization with overlap
> >   ethr =  1.00E-05,  avg # of iterations = 14.0
> >
>  
> %%
> >   Error in routine tau_xc (5):
> >
>  
> %%
> >
>  
> %%
> >
>  
> %%
> >   This case is not implemented
> >
>  
> %%
> >   Error in routine tau_xc (5):
> >   This case is not implemented
> >   Error in routine tau_xc (5):
> > Am I missing something? It is simple "scf" calculation
> > with ONCVPSP pseudopotentials.I even tried SCAN pseudo
> > without 'input_dft' option and still got the same error.
> > Thanks,
> > Sergey
> >   University of South Florida.
> > ___
> > Quantum ESPRESSO is supported by MaX
> >

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Lorenzo Paulatto 

Your suggestion (input_dft = 'MGGA_X_SCAN MGGA_C_SCAN') worked.



"Got the code to spit out a number" is not the same as "produced a valid 
result". I would wait for an answer from someone knowledgeable on this 
subject before spending too much CPU  time in this case!


Or even better open an issue in the gitlab tracker 
 explaining why do you think that 
two equivalent cases are producing different outcomes. This was you'll 
have an answer for sure


cheers

However, I still don't understand why others could use  (input_dft = 
'scan') without any problem - I searched archive and saw several 
examples. Also, I used SCAN pseudopotentials from QE website, so in 
principle I shouldn't declare input_dft option. And not declaring would 
result in error (as I reported).

Thanks,
Sergey Lisenkov
26.05.2020, 10:36, "Fabrizio Ferrari" :

Hello Sergey,
that is strange. I have tried some simple tests too, but I could not
reproduce your issue. An equivalent way to activate the scan mgga
functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'
You can try with that, but if it still doesn't work, I think I would
need more details about your input in order to find the issue.
Cheers
On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov mailto:proff...@yandex.ru>> wrote:

Hello Fabrizio,
yes, it is compiled with libxc support. Without it, pw.x
complains immediately, that it should be compiled with Libxc
library (or something like that).
25.05.2020, 17:21, "Fabrizio Ferrari"
mailto:ferrariruffino...@gmail.com>>:

Hello,
in your case (scan) that error should come out only if libxc
has not been properly linked during compilation. Have you
tried runs with other libxc functionals (not necessarily
mGGA)? Do they work?
Just in case you want to check, to properly compile with
libxc, after the qe configuration, you have to add the
requested flags in the make.inc file at 'DFLAGS', 'IFLAGS'
and 'LD_LIBS' (you can follow the nearby comments).
Fabrizio
On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov
mailto:proff...@yandex.ru>> wrote:

Hello,
I noticed that some people run SCAN calculations with QE
v.6.5. I tried this, but my test has failed.
I compiled 6.5 version with LibXC-4.3.4. When I activate
"input_dft = 'scan'" in my input file,
I get the following error just before first scf step:

      total cpu time spent up to now is        1.0 secs
      Self-consistent Calculation
      iteration #  1     ecut=   100.00 Ry     beta= 0.70
      Davidson diagonalization with overlap
      ethr =  1.00E-05,  avg # of iterations = 14.0
  
%%
      Error in routine tau_xc (5):
  
%%
  
%%
  
%%
      This case is not implemented
  
%%
      Error in routine tau_xc (5):
      This case is not implemented
      Error in routine tau_xc (5):
Am I missing something? It is simple "scf" calculation
with ONCVPSP pseudopotentials.I even tried SCAN pseudo
without 'input_dft' option and still got the same error.
Thanks,
Sergey
  University of South Florida.
___
Quantum ESPRESSO is supported by MaX
(www.max-centre.eu/quantum-espresso
)
users mailing list users@lists.quantum-espresso.org

https://lists.quantum-espresso.org/mailman/listinfo/users

,

___
Quantum ESPRESSO is supported by MaX
(www.max-centre.eu/quantum-espresso
)
users mailing list users@lists.quantum-espresso.org

https://lists.quantum-espresso.org/mailman/listinfo/users

___
Quantum ESPRESSO is

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Fabrizio Ferrari 
I don't understand too. I have tried a couple of tests both with 'SCAN' and
with the explicit form 'MGGA_X_SCAN' and they both worked in the same
way.
If your code has not been modified, there must be something very specific
to your input that causes the bug. I'll try to investigate more.
Cheers,
Fabrizio

On Tue, May 26, 2020 at 11:18 AM Sergey Lisenkov  wrote:

> Hello Fabrizio,
>
> Your suggestion (input_dft = 'MGGA_X_SCAN MGGA_C_SCAN') worked.
>
> However, I still don't understand why others could use  (input_dft =
> 'scan') without any problem - I searched archive and saw several examples.
> Also, I used SCAN pseudopotentials from QE website, so in principle I
> shouldn't declare input_dft option. And not declaring would result in error
> (as I reported).
>
> Thanks,
> Sergey Lisenkov
>
> 26.05.2020, 10:36, "Fabrizio Ferrari" :
>
> Hello Sergey,
> that is strange. I have tried some simple tests too, but I could not
> reproduce your issue. An equivalent way to activate the scan mgga
> functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'
> You can try with that, but if it still doesn't work, I think I would need
> more details about your input in order to find the issue.
> Cheers
>
> On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov 
> wrote:
>
> Hello Fabrizio,
>
> yes, it is compiled with libxc support. Without it, pw.x complains
> immediately, that it should be compiled with Libxc library (or something
> like that).
>
>
> 25.05.2020, 17:21, "Fabrizio Ferrari" :
>
> Hello,
> in your case (scan) that error should come out only if libxc has not been
> properly linked during compilation. Have you tried runs with other libxc
> functionals (not necessarily mGGA)? Do they work?
> Just in case you want to check, to properly compile with libxc, after the
> qe configuration, you have to add the requested flags in the make.inc file
> at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).
> Fabrizio
>
>
> On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov 
> wrote:
>
> Hello,
>
> I noticed that some people run SCAN calculations with QE v.6.5. I tried
> this, but my test has failed.
>
> I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft =
> 'scan'" in my input file,
>
> I get the following error just before first scf step:
>
> 
>
>  total cpu time spent up to now is1.0 secs
>
>  Self-consistent Calculation
>
>  iteration #  1 ecut=   100.00 Ry beta= 0.70
>  Davidson diagonalization with overlap
>  ethr =  1.00E-05,  avg # of iterations = 14.0
>
>
>  
> %%
>
>  Error in routine tau_xc (5):
>
>
>  
> %%
>
>  
> %%
>
>
>
>  
> %%
>
>  This case is not implemented
>
>
>
>
>  
> %%
>  Error in routine tau_xc (5):
>  This case is not implemented
>  Error in routine tau_xc (5):
>
> Am I missing something? It is simple "scf" calculation with ONCVPSP
> pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and
> still got the same error.
>
> Thanks,
> Sergey
>  University of South Florida.
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ,
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ,
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Sergey Lisenkov 
Hello Fabrizio, Your suggestion (input_dft = 'MGGA_X_SCAN MGGA_C_SCAN') worked. However, I still don't understand why others could use  (input_dft = 'scan') without any problem - I searched archive and saw several examples. Also, I used SCAN pseudopotentials from QE website, so in principle I shouldn't declare input_dft option. And not declaring would result in error (as I reported). Thanks,Sergey Lisenkov 26.05.2020, 10:36, "Fabrizio Ferrari" :Hello Sergey,that is strange. I have tried some simple tests too, but I could not reproduce your issue. An equivalent way to activate the scan mgga functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'You can try with that, but if it still doesn't work, I think I would need more details about your input in order to find the issue.Cheers On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov  wrote:Hello Fabrizio, yes, it is compiled with libxc support. Without it, pw.x complains immediately, that it should be compiled with Libxc library (or something like that).  25.05.2020, 17:21, "Fabrizio Ferrari" :Hello,in your case (scan) that error should come out only if libxc has not been properly linked during compilation. Have you tried runs with other libxc functionals (not necessarily mGGA)? Do they work?Just in case you want to check, to properly compile with libxc, after the qe configuration, you have to add the requested flags in the make.inc file at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).Fabrizio  On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov  wrote:Hello, I noticed that some people run SCAN calculations with QE v.6.5. I tried this, but my test has failed. I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft = 'scan'" in my input file, I get the following error just before first scf step:       total cpu time spent up to now is        1.0 secs      Self-consistent Calculation      iteration #  1     ecut=   100.00 Ry     beta= 0.70     Davidson diagonalization with overlap     ethr =  1.00E-05,  avg # of iterations = 14.0  %%      Error in routine tau_xc (5):  %% %%   %%      This case is not implemented%%     Error in routine tau_xc (5):     This case is not implemented     Error in routine tau_xc (5): Am I missing something? It is simple "scf" calculation with ONCVPSP pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and still got the same error. Thanks,Sergey University of South Florida.___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users,___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users,___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Sergey Lisenkov 
Hello Fabrizio, Your suggestion 26.05.2020, 10:36, "Fabrizio Ferrari" :Hello Sergey,that is strange. I have tried some simple tests too, but I could not reproduce your issue. An equivalent way to activate the scan mgga functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'You can try with that, but if it still doesn't work, I think I would need more details about your input in order to find the issue.Cheers On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov  wrote:Hello Fabrizio, yes, it is compiled with libxc support. Without it, pw.x complains immediately, that it should be compiled with Libxc library (or something like that).  25.05.2020, 17:21, "Fabrizio Ferrari" :Hello,in your case (scan) that error should come out only if libxc has not been properly linked during compilation. Have you tried runs with other libxc functionals (not necessarily mGGA)? Do they work?Just in case you want to check, to properly compile with libxc, after the qe configuration, you have to add the requested flags in the make.inc file at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).Fabrizio  On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov  wrote:Hello, I noticed that some people run SCAN calculations with QE v.6.5. I tried this, but my test has failed. I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft = 'scan'" in my input file, I get the following error just before first scf step:       total cpu time spent up to now is        1.0 secs      Self-consistent Calculation      iteration #  1     ecut=   100.00 Ry     beta= 0.70     Davidson diagonalization with overlap     ethr =  1.00E-05,  avg # of iterations = 14.0  %%      Error in routine tau_xc (5):  %% %%   %%      This case is not implemented%%     Error in routine tau_xc (5):     This case is not implemented     Error in routine tau_xc (5): Am I missing something? It is simple "scf" calculation with ONCVPSP pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and still got the same error. Thanks,Sergey University of South Florida.___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users,___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users,___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-26 Thread Fabrizio Ferrari 
Hello Sergey,
that is strange. I have tried some simple tests too, but I could not
reproduce your issue. An equivalent way to activate the scan mgga
functionals is to put:  input_dft = 'MGGA_X_SCAN MGGA_C_SCAN'
You can try with that, but if it still doesn't work, I think I would need
more details about your input in order to find the issue.
Cheers

On Mon, May 25, 2020 at 8:16 PM Sergey Lisenkov  wrote:

> Hello Fabrizio,
>
> yes, it is compiled with libxc support. Without it, pw.x complains
> immediately, that it should be compiled with Libxc library (or something
> like that).
>
>
> 25.05.2020, 17:21, "Fabrizio Ferrari" :
>
> Hello,
> in your case (scan) that error should come out only if libxc has not been
> properly linked during compilation. Have you tried runs with other libxc
> functionals (not necessarily mGGA)? Do they work?
> Just in case you want to check, to properly compile with libxc, after the
> qe configuration, you have to add the requested flags in the make.inc file
> at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).
> Fabrizio
>
>
> On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov 
> wrote:
>
> Hello,
>
> I noticed that some people run SCAN calculations with QE v.6.5. I tried
> this, but my test has failed.
>
> I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft =
> 'scan'" in my input file,
>
> I get the following error just before first scf step:
>
> 
>
>  total cpu time spent up to now is1.0 secs
>
>  Self-consistent Calculation
>
>  iteration #  1 ecut=   100.00 Ry beta= 0.70
>  Davidson diagonalization with overlap
>  ethr =  1.00E-05,  avg # of iterations = 14.0
>
>
>  
> %%
>
>  Error in routine tau_xc (5):
>
>
>  
> %%
>
>  
> %%
>
>
>
>  
> %%
>
>  This case is not implemented
>
>
>
>
>  
> %%
>  Error in routine tau_xc (5):
>  This case is not implemented
>  Error in routine tau_xc (5):
>
> Am I missing something? It is simple "scf" calculation with ONCVPSP
> pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and
> still got the same error.
>
> Thanks,
> Sergey
>  University of South Florida.
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ,
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-25 Thread Sergey Lisenkov 
Hello Fabrizio, yes, it is compiled with libxc support. Without it, pw.x complains immediately, that it should be compiled with Libxc library (or something like that).25.05.2020, 17:21, "Fabrizio Ferrari" :Hello,in your case (scan) that error should come out only if libxc has not been properly linked during compilation. Have you tried runs with other libxc functionals (not necessarily mGGA)? Do they work?Just in case you want to check, to properly compile with libxc, after the qe configuration, you have to add the requested flags in the make.inc file at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).FabrizioOn Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov  wrote:Hello, I noticed that some people run SCAN calculations with QE v.6.5. I tried this, but my test has failed. I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft = 'scan'" in my input file, I get the following error just before first scf step:       total cpu time spent up to now is        1.0 secs      Self-consistent Calculation      iteration #  1     ecut=   100.00 Ry     beta= 0.70     Davidson diagonalization with overlap     ethr =  1.00E-05,  avg # of iterations = 14.0  %%      Error in routine tau_xc (5):  %% %%   %%      This case is not implemented%%     Error in routine tau_xc (5):     This case is not implemented     Error in routine tau_xc (5): Am I missing something? It is simple "scf" calculation with ONCVPSP pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and still got the same error. Thanks,Sergey University of South Florida.___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users
,___Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)users mailing list users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] SCAN functional in QE-6.5

2020-05-25 Thread Fabrizio Ferrari 
Hello,
in your case (scan) that error should come out only if libxc has not been
properly linked during compilation. Have you tried runs with other libxc
functionals (not necessarily mGGA)? Do they work?
Just in case you want to check, to properly compile with libxc, after the
qe configuration, you have to add the requested flags in the make.inc file
at 'DFLAGS', 'IFLAGS' and 'LD_LIBS' (you can follow the nearby comments).
Fabrizio


On Mon, May 25, 2020 at 1:28 PM Sergey Lisenkov  wrote:

> Hello,
>
> I noticed that some people run SCAN calculations with QE v.6.5. I tried
> this, but my test has failed.
>
> I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft =
> 'scan'" in my input file,
>
> I get the following error just before first scf step:
>
> 
>
>  total cpu time spent up to now is1.0 secs
>
>  Self-consistent Calculation
>
>  iteration #  1 ecut=   100.00 Ry beta= 0.70
>  Davidson diagonalization with overlap
>  ethr =  1.00E-05,  avg # of iterations = 14.0
>
>
>  
> %%
>
>  Error in routine tau_xc (5):
>
>
>  
> %%
>
>  
> %%
>
>
>
>  
> %%
>
>  This case is not implemented
>
>
>
>
>  
> %%
>  Error in routine tau_xc (5):
>  This case is not implemented
>  Error in routine tau_xc (5):
>
> Am I missing something? It is simple "scf" calculation with ONCVPSP
> pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and
> still got the same error.
>
> Thanks,
> Sergey
>  University of South Florida.
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

[QE-users] SCAN functional in QE-6.5

2020-05-25 Thread Sergey Lisenkov 
Hello, I noticed that some people run SCAN calculations with QE v.6.5. I tried this, but my test has failed. I compiled 6.5 version with LibXC-4.3.4. When I activate "input_dft = 'scan'" in my input file, I get the following error just before first scf step:       total cpu time spent up to now is        1.0 secs      Self-consistent Calculation      iteration #  1     ecut=   100.00 Ry     beta= 0.70     Davidson diagonalization with overlap     ethr =  1.00E-05,  avg # of iterations = 14.0  %%      Error in routine tau_xc (5):  %% %%   %%      This case is not implemented%%     Error in routine tau_xc (5):     This case is not implemented     Error in routine tau_xc (5): Am I missing something? It is simple "scf" calculation with ONCVPSP pseudopotentials.I even tried SCAN pseudo without 'input_dft' option and still got the same error. Thanks,Sergey University of South Florida.___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users