Error In Routine Mix_rho
(1): negative dr2 Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] Dear colleagues, I am reporting an error that I have been experiencing with quantum_espresso-5.0.x (x=1,2). Running a 'relax' calculation in a CdSe wurtzite structure with the following pseudopotenciales: Cd.pbe-dn-kjpaw_psl.0.2.UPF Se.pbe-n-kjpaw_psl.0.2.UPF I have this error: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Error in routine mix_rho (1): negative dr2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% However, if I run the same 'relax' calculation with the following pseudopotentials: Cd.pbe-dn-rrkjus_psl.0.2.UPF Se.pbe-n-rrkjus_psl.0.2.UPF I do not have the above error. All pseudopotentials were downloaded from Quantum Espresso website. Below is the code that I have been using for 'relax' calculation with both sets of pseudopotentials. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% &CONTROL calculation = 'relax' , restart_mode = 'from_scratch' , outdir = './tmp' , pseudo_dir = './' , prefix = 'cdse' , etot_conv_thr = 1.0d-6 , forc_conv_thr = 1.0d-4 , tstress = .true. , tprnfor = .true. , / &SYSTEM ibrav = 0, celldm(1) = 1.88972613289, nat = 4, ntyp = 2, ecutwfc = 40.0 , ecutrho = 250.0 , exxdiv_treatment = 'gygi-baldereschi' , / &ELECTRONS conv_thr = 1.0d-8 , / &IONS ion_dynamics = 'bfgs' , / CELL_PARAMETERS alat 4.382143000 0.000000000 0.000000000 -2.191071500 3.795047161 0.000000000 -0.000000000 -0.000000000 7.149948000 ATOMIC_SPECIES Cd 112.40000 Cd.pbe-dn-kjpaw_psl.0.2.UPF Se 78.96000 Se.pbe-n-kjpaw_psl.0.2.UPF ATOMIC_POSITIONS crystal Cd 0.666667000 0.333333000 0.000000000 Se 0.666667000 0.333333000 0.375000000 Cd 0.333333000 0.666667000 0.500000000 Se 0.333333000 0.666667000 0.875000000 K_POINTS automatic 9 9 6 0 0 0 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Running the same kind of calculation with quantum_espresso-4.3 I have not experienced this error. Sincerely Yours Daniel Cesar ------------------------------------------------------------ Daniel Cesar, Post Doctoral Researcher*, *Dept of Physics, Universidad de Chile email. dfcesar at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pwscf.org/pipermail/pw_forum/attachments/20130507/4a62ceea/attachment.html Previous message: [Pw_forum] Performance problems due to files in outdir Next message: [Pw_forum] Error in routine mix_rho (1): negative dr2 Messages sorted by: [ date ] [ thread ] [ subject ] [ aut
> > Running the same kind of calculation with quantum_espresso-4.3 I have > not experienced this error. because the calculation of the error estimate with PAW was slightly http://qe-forge.org/pipermail/pw_forum/2013-May/101723.html different in 4.3.2. After the error estimate was improved, occasionally one finds that the error estimate, a positive-definite quantity, becomes negative. In the svn version we reverted to the old estimate, while waiting https://www.mail-archive.com/pw_forum@pwscf.org/msg20612.html for better ideas. P. -- Paolo Giannozzi, Dept. Chemistry&Physics&Environment, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222 Previous message View by thread View by date Next message [Pw_forum] Error in routine mix_rho (1): negative dr2 Daniel Cesar [Pw_forum] Error in routine mix_rho (1): negative dr2 Yue-Wen Fang [Pw_forum] Error in routine mix_rho (1): negative dr2 Paolo Giannozzi Reply via email to Search the site The Mail Archive home pw_forum - all messages pw_forum - about the list Expand Previous message Next message The Mail Archive home Add your mailing list FAQ Support Privacy 1368121590.6321.3.camel@pania.fastwebnet.it
Sign in Pricing Blog Support Search GitHub This repository Watch 7 Star 13 Fork 19 NNemec/quantum-espresso Code Issues 1 Pull requests 0 Projects https://github.com/NNemec/quantum-espresso/blob/master/PW/mix_rho.f90 0 Pulse Graphs Permalink Branch: master Switch branches/tags Branches Tags O-sesame QE-3-0 QE-4-0 QE-4-1 QE-4-2 develop_PAW master obm origin Nothing to show Nothing to show Find file Copy path quantum-espresso/PW/mix_rho.f90 https://trac-foundry.lbl.gov/trac/export/321/ESPRESSO/branches/shirley_QE4.3/PW/electrons.f90 Fetching contributors… Cannot retrieve contributors at this time Raw Blame History 677 lines (671 sloc) 18.4 KB ! ! Copyright (C) 2002-2007 Quantum ESPRESSO group ! This file is error in distributed under the terms of the ! GNU General Public License. See the file `License' ! in the root directory of the present distribution, ! or http://www.gnu.org/copyleft/gpl.txt . ! ! #define ZERO ( 0._dp, 0._dp ) ! ! This macro force the normalization of betamix matrix, usually not necessary !#define __NORMALIZE_BETAMIX ! #ifdef __GFORTRAN ! gfortran hack - for some error in routine mysterious reason gfortran doesn't save ! derived-type variables even with the SAVE attribute MODULE mix_save USE scf, ONLY : mix_type TYPE(mix_type), ALLOCATABLE, SAVE :: & df(:), &! information from preceding iterations dv(:) ! " " " " " " END MODULE mix_save #endif !---------------------------------------------------------------------------- SUBROUTINE mix_rho( input_rhout, rhoin, alphamix, dr2, tr2_min, iter, n_iter, conv ) !---------------------------------------------------------------------------- ! ! ... Modified Broyden's method for charge density mixing ! ... D.D. Johnson PRB 38, 12807 (1988) ! ! ... On output: the mixed density is in rhoin, mixed augmentation ! ... channel occ. is in becin ! input_rhocout, input_becout etc are unchanged ! USE kinds, ONLY : DP USE ions_base, ONLY : nat USE gvect, ONLY : ngm USE gsmooth, ONLY : ngms USE lsda_mod, ONLY : nspin USE control_flags, ONLY : imix, ngm0, tr2, io_level USE io_files, ONLY : find_free_unit ! ... for PAW: USE uspp_param, ONLY : nhm USE scf, ONLY : scf_type, create_scf_type, destroy_scf_type, & mix_type, create_mix_type, destroy_mix_type, & assign_scf_to_mix_type, assign_mix_to_scf_type, & mix_type_AXPY, diropn_mix_file, close_mix_file, & davcio_mix_type, rho_ddot, high_frequency_mixing, & mix_type_COPY, mix_type_SCAL USE io_global, O