In case there is any confusion, there was a typo in my last statement; inverted
commas in the wrong place. I should have said
Remove the "input_dft = 'read from pseudopotential files' " from &system.
i.e. your system namelist should look like
&system
ibrav = 0
celldm(1) = 12.0
nat = 5
ntyp = 3
nbnd = 21
ecutwfc = 30.0
ecutrho = 360.0
occupations = 'fixed'
/
...
________________________________
From: [email protected] [[email protected]] on behalf of
Gabriel Greene [[email protected]]
Sent: Monday, April 27, 2015 1:44 PM
To: PWSCF Forum
Subject: Re: [Pw_forum] Unrecognized dft error
Hi,
The error is syntactical, nothing to do with your pseudopotentials. Remove the
input_dft = " 'read from pseudopotential files' " from &system. Unless you want
to use hybrid DFT funcitonals (see input data description on the QE website),
you don't need to specify input_dft.
Gabriel Greene,
Electronics Theory Group,
Tyndall National Institute,
Cork, Ireland
________________________________
From: [email protected] [[email protected]] on behalf of Eric
Glen Suter [[email protected]]
Sent: Monday, April 27, 2015 12:55 PM
To: [email protected]
Subject: [Pw_forum] Unrecognized dft error
Hello all,
I'm currently trying to study KNbO3 with quantum espresso version 5.1.2. The
relevant parts of my input files look like this:
&system
ibrav = 0
celldm(1) = 12.0
nat = 5
ntyp = 3
nbnd = 21
ecutwfc = 30.0
ecutrho = 360.0
occupations = 'fixed'
input_dft = 'read from pseudopotential files'
/
&electrons
electron_maxstep = 1000
conv_thr = 1e-8,
mixing_mode = 'plain'
mixing_beta = 0.7,
diagonalization = 'david'
startingpot = 'atomic'
startingwfc = 'atomic+random'
/
ATOMIC_SPECIES
K 3.90983000E+01 K.pz-sp-van.UPF
Nb 9.29063800E+01 Nb.pbe-nsp-van.UPF
O 1.59994000E+01 O.pbe-van_ak.UPF
However, when I try to run pw.x with this input file, I get an error that reads:
Error in routine set_dft_from_name (1):
READ FROM PSEUDOPOTENTIAL FILES: unrecognized dft
The pseudopotentials I have referenced all either came with the initial build
of 5.1.2 or were downloaded from the quantum-espresso site. After checking
them, it looks as though they should all have valid strings for funct.f90 to
identify them.
K.pz-sp-van has: SLA PZ NOGX NOGC
Nb.pbe-nsp-van has: SLA PW PBE PBE
and O.pbe-van_ak has: SLA PZ NOGX NOGC
Is there something incompatible with the functionals they're specifying or is
there something else that I'm missing? Any insight would be tremendously
appreciated. Thank you for your time.
Best regards,
Eric Suter
University of Georgia
Physics Department
_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum
