Skip to content

GitLab

fleur
fleur
Commit e18fe3aa authored by Uliana Alekseeva's avatar Uliana Alekseeva
Browse files
Options

Merge branch 'develop' of ifflinux.iff.kfa-juelich.de:fleur into develop

parents 8fd6a2b7 4eaafae1
Hide whitespace changes
Inline Side-by-side
Showing with 631 additions and 84 deletions
.gitlab-ci.yml
View file @ e18fe3aa
stages:
- build
- test
- deploy
- build-pgi
- test-pgi
build-gfortran:
image: ubuntu:16.10
image: iffregistry.fz-juelich.de/fleur/fleur:latest
stage: build
cache:
paths:
- build
script:
- apt-get update -qq && apt-get install -y -qq build-essential gfortran gfortran-6 g++-6 gcc-6 binutils-gold git cmake libxml2-dev liblapack-dev libblas-dev doxygen libopenmpi-dev libscalapack-mpi-dev
- cd /builds/fleur/fleur; ./configure.sh GITLAB; cd build; make
test-gfortran:
image: ubuntu:16.10
image: iffregistry.fz-juelich.de/fleur/fleur:latest
stage: test
cache:
paths:
- build
script:
- apt-get update -qq && apt-get install -y -qq build-essential gfortran gfortran-6 g++-6 gcc-6 binutils-gold git cmake libxml2-dev liblapack-dev libblas-dev libopenmpi-dev libscalapack-mpi-dev
- ulimit -s unlimited ;export juDFT_MPI="mpirun -n 2 --allow-run-as-root ";cd /builds/fleur/fleur/build;ctest
pages:
image: iffregistry.fz-juelich.de/fleur/fleur:latest
stage: deploy
cache:
paths:
- build
script:
- cd /builds/fleur/fleur/build ; make doc
- mv docs/html/ ../public/
artifacts:
paths:
- public
- build/fleur
- build/fleur_MPI
- build/inpgen
#build-pgi:
# image: hopobcn/pgi-ce
# stage: build
# cache:
# paths:
# - build.debug
# script:
# - apt-get update -qq && apt-get install -y -qq cmake libxml2-dev
# - cd /builds/fleur/fleur; ./configure.sh AUTO debug; cd build.debug; make
#test-pgi:
# image: hopobcn/pgi-ce
# stage: test
# cache:
# paths:
# - build.debug
# script:
# - apt-get update -qq && apt-get install -y -qq cmake libxml2-dev
# - cd /builds/fleur/fleur/build.debug;ctest
build-pgi:
image: iffregistry.fz-juelich.de/fleur/fleur:pgi
stage: build-pgi
cache:
paths:
- build.debug
script:
- apt-get update -qq && apt-get install -y -qq cmake libxml2-dev
- cd /builds/fleur/fleur; ./configure.sh AUTO debug; cd build.debug; make
only:
- schedules
- triggers
- web
test-pgi:
image: iffregistry.fz-juelich.de/fleur/fleur:pgi
stage: test-pgi
cache:
paths:
- build.debug
script:
- apt-get update -qq && apt-get install -y -qq cmake libxml2-dev
- cd /builds/fleur/fleur/build.debug;ctest
only:
- schedules
- web
- triggers
docs/ContributorsGuide.md 0 → 100644
View file @ e18fe3aa
Contributors guide
======================================
Everyone is very welcome to contribute to the enhancement of FLEUR.
Please use the [gitlab service] (https://iffgit.fz-juelich.de/fleur/fleur) to obtain the
latest development version of FLEUR.
##Coding rules for FLEUR:
In no particular order we try to collect items to consider when writing code for FLEUR
- Instead of 'stop' use calls to judft_error, judft_warn, judft_end
- Do not read and write any files. Files are neither replacements for common-blocks nor a
storage for status variables.
Only exceptions:
+ you create nice IO subroutines in the subdirectory io
+ you write to the typical FLEUR output files
docs/Main.md 0 → 100644
View file @ e18fe3aa
\mainpage Welcome to the developers documentation of FLEUR
These pages are complementing the <A HREF="http://www.flapw.de/pm/index.php?n=User-Documentation.Documentation"> User documentation </A> in which the actual use of FLEUR is documented.
Here we collect information useful for developers or advanced users that actually will modify the source code.
## GitLab
The development process is performed using gitlab. You can access the <A HREF="https://iffgit.fz-juelich.de/fleur/fleur"> main gitlab page here.</A>
If you checkout the code please be aware that there are several branches. The release branch contains the code of the last release published on <A HREF="http://www.flapw.de/pm/index.php?n=FLEUR.Downloads"> the FLEUR webpage</A>. You can *not* push to this branch directly.
You probably want to use the development branch to insert your changes. If your changes are large, it might be a good idea to create your own branch first.
## More info
In the menu to the left you will find information on various topics as well as the links to the source-code documentation.
docs/Tips+Tricks.md 0 → 100644
View file @ e18fe3aa
Useful info for developers
==============================================
## Using fleur with the HDF5 library and debugging it with valgrind
HDF5 has to be built with the same compiler that is also used to compile fleur. If adapted the following commands can be used to compile a HDF5 library for fleur:
+ 'FC=/usr/local/intel/impi/4.0.3.008/intel64/bin/mpiifort CC=/usr/local/intel/impi/4.0.3.008/intel64/bin/mpiicc CXX=/usr/local/intel/impi/4.0.3.008/intel64/bin/mpiicc ./configure --enable-fortran --enable-fortran2003 --enable-parallel --enable-using-memchecker --enable-clear-file-buffers'
+ 'make'
+ 'make install'
+ 'make check' (optional)
Note:
+ The paths have to be adjusted such that that compiler is used which is also used to compile fleur.
+ The parallel switch is not needed for every calculation: Only for parallel calculations in which HDF5 is also used for the eigenvector IO.
+ The last two command line switches in the configure command turn on initializations of irrelevant array parts in HDF5. If valgrind is not needed it is probably the better choice to leave them away. If left away valgrind will complain about missing initializations in the HDF5 library.
+ valgrind gives partially strange behavior if used together with the intel compiler. It would be better to use it together with gfortran.
+ At the moment HDF5 is needed in version 1.8.*. Usage of version 1.10.* yields some problems.
Furthermore to configure and start fleur with HDF5 the following has to be done:
+ In your .bashrc the HDF5 library has to be added to the LD_LIBRARY_PATH. This implies a line like 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/hdf5/current/hdf5/lib'
+ configure fleur with some line like 'CMAKE_Fortran_FLAGS="-I~/hdf5/current/hdf5/include" FLEUR_LIBRARIES="-L~/hdf5/current/hdf5/lib;-lhdf5_fortran;-lhdf5" ./fleur/configure.sh IFF'
\ No newline at end of file
eigen/eigen.F90
View file @ e18fe3aa
......@@ -17,7 +17,7 @@ CONTAINS
! nmat, nbasfcn including LO's
! g. bihlmayer '96
!**********************************************************************
USE m_constants, ONLY : pi_const,sfp_const
USE m_constants
USE m_types
USE m_lodpot
USE m_tlmplm
......@@ -75,7 +75,7 @@ CONTAINS
!+odim
! ..
! .. Local Scalars ..
INTEGER jsp,nk,nred,ne_all,n_u_in,ne_found
INTEGER jsp,nk,nred,ne_all,ne_found
INTEGER ne,matsize ,nrec,lh0
INTEGER nspins,isp,i,j,err
INTEGER mlotot,mlolotot
......@@ -83,13 +83,11 @@ CONTAINS
! ..
! .. Local Arrays ..
INTEGER, PARAMETER :: lmaxb=3
INTEGER, ALLOCATABLE :: matind(:,:),kveclo(:)
INTEGER, ALLOCATABLE :: nv2(:)
REAL, ALLOCATABLE :: bkpt(:)
REAL, ALLOCATABLE :: eig(:)
COMPLEX, ALLOCATABLE :: vs_mmp(:,:,:,:)
TYPE(t_tlmplm) :: td
TYPE(t_usdus) :: ud
TYPE(t_lapw) :: lapw
......@@ -203,7 +201,10 @@ CONTAINS
matsize = ((DIMENSION%nbasfcn+1)/2)*DIMENSION%nbasfcn
ENDIF
ENDIF
IF (matsize<2) CALL judft_error("Wrong size of matrix",calledby="eigen",hint="Your basis might be too large or the parallelization fail or ??")
IF (matsize<2) THEN
CALL judft_error("Wrong size of matrix",calledby="eigen",&
hint="Your basis might be too large or the parallelization fail or ??")
END IF
ne = DIMENSION%neigd
eig_id=open_eig(&
......@@ -247,12 +248,11 @@ CONTAINS
!
! ..
! LDA+U
n_u_in=atoms%n_u
IF ((atoms%n_u.GT.0)) THEN
ALLOCATE( vs_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,input%jspins) )
CALL u_setup(sym,atoms,lmaxb,sphhar,input, enpara%el0(0:,:,:),v%mt,mpi, vs_mmp,results)
ALLOCATE( v%mmpMat(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u,input%jspins) )
CALL u_setup(sym,atoms,sphhar,input, enpara%el0(0:,:,:),v,mpi,results)
ELSE
ALLOCATE( vs_mmp(-lmaxb:-lmaxb,-lmaxb:-lmaxb,1,2) )
ALLOCATE( v%mmpMat(-lmaxU_const:-lmaxU_const,-lmaxU_const:-lmaxU_const,1,2) )
ENDIF
!
!---> loop over k-points: each can be a separate task
......@@ -282,7 +282,7 @@ CONTAINS
ENDIF
lh0=1
CALL tlmplm(sphhar,atoms,DIMENSION,enpara, jsp,1,mpi, v%mt(1,0,1,jsp),lh0,input, td,ud)
IF (input%l_f) CALL write_tlmplm(td,vs_mmp,atoms%n_u>0,1,jsp,input%jspins)
IF (input%l_f) CALL write_tlmplm(td,v%mmpMat,atoms%n_u>0,1,jsp,input%jspins)
CALL timestop("tlmplm")
!---> pk non-collinear
......@@ -293,7 +293,7 @@ CONTAINS
isp = 2
CALL timestart("tlmplm")
CALL tlmplm(sphhar,atoms,DIMENSION,enpara,isp,isp,mpi, v%mt(1,0,1,isp),lh0,input, td,ud)
IF (input%l_f) CALL write_tlmplm(td,vs_mmp,atoms%n_u>0,2,2,input%jspins)
IF (input%l_f) CALL write_tlmplm(td,v%mmpMat,atoms%n_u>0,2,2,input%jspins)
CALL timestop("tlmplm")
ENDIF
!
......@@ -332,7 +332,7 @@ CONTAINS
IF (.NOT.l_wu) THEN
CALL timestart("MT Hamiltonian&Overlap")
CALL hsmt(DIMENSION,atoms,sphhar,sym,enpara, mpi%SUB_COMM,mpi%n_size,mpi%n_rank,jsp,input,mpi,&
lmaxb, noco,cell, lapw, bkpt,v%mt, vs_mmp, oneD,ud, kveclo,td,l_real,hamOvlp)
lmaxU_const, noco,cell, lapw, bkpt,v%mt,v%mmpMat, oneD,ud, kveclo,td,l_real,hamOvlp)
CALL timestop("MT Hamiltonian&Overlap")
ENDIF
!
......@@ -423,7 +423,7 @@ ENDIF
DEALLOCATE (td%ind,td%tuulo,td%tdulo)
DEALLOCATE (td%tuloulo)
END DO ! spin loop ends
DEALLOCATE( vs_mmp )
DEALLOCATE(v%mmpMat)
DEALLOCATE (matind)
IF (l_real) THEN
DEALLOCATE(hamOvlp%a_r,hamOvlp%b_r)
......@@ -439,8 +439,6 @@ ENDIF
CALL MPI_BARRIER(mpi%MPI_COMM,ierr)
#endif
!IF (hybrid%l_hybrid.OR.hybrid%l_calhf) CALL close_eig(eig_id)
atoms%n_u=n_u_in
IF( input%jspins .EQ. 1 .AND. hybrid%l_hybrid ) THEN
results%te_hfex%valence = 2*results%te_hfex%valence
......
force/force_a21.F90
View file @ e18fe3aa
......@@ -54,7 +54,6 @@ CONTAINS
COMPLEX, INTENT (IN) :: cveccof(3,-atoms%llod:atoms%llod,nobd,atoms%nlod,atoms%nat)
! ..
! .. Local Scalars ..
INTEGER, PARAMETER :: lmaxb=3
COMPLEX dtd,dtu,utd,utu
INTEGER lo, mlotot, mlolotot, mlot_d, mlolot_d
INTEGER i,ie,im,in,l1,l2,ll1,ll2,lm1,lm2,m1,m2,n,natom,m,i_u
......@@ -88,7 +87,7 @@ CONTAINS
a21(3,atoms%nat),b4(3,atoms%nat),tlmplm%ind(0:DIMENSION%lmd,0:DIMENSION%lmd,atoms%ntype,1) )
!
IF(atoms%n_u.GT.0) THEN
ALLOCATE(v_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u))
ALLOCATE(v_mmp(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u))
v_mmp = CMPLX(0.0,0.0)
CALL read_tlmplm_vs_mmp(jsp,atoms%n_u,v_mmp)
END IF
......@@ -197,7 +196,7 @@ CONTAINS
cveccof, tlmplm,usdus, a21)
IF ((atoms%n_u.GT.0).AND.(i_u.LE.atoms%n_u)) THEN
CALL force_a21_U(nobd,atoms,lmaxb,i_u,n,jsp,we,ne,&
CALL force_a21_U(nobd,atoms,i_u,n,jsp,we,ne,&
usdus,v_mmp,acof,bcof,ccof,&
aveccof,bveccof,cveccof, a21)
END IF
......
force/force_a21_U.f90
View file @ e18fe3aa
MODULE m_forcea21U
CONTAINS
SUBROUTINE force_a21_U(nobd,atoms,lmaxb,i_u,itype,isp,we,ne,&
SUBROUTINE force_a21_U(nobd,atoms,i_u,itype,isp,we,ne,&
usdus,v_mmp, acof,bcof,ccof,aveccof,bveccof,cveccof, a21)
!
!***********************************************************************
......@@ -9,6 +9,7 @@ CONTAINS
! Comp.Phys.Comm. 179 (2008) 784-790
!***********************************************************************
!
USE m_constants
USE m_types
IMPLICIT NONE
......@@ -17,13 +18,13 @@ CONTAINS
! ..
! .. Scalar Arguments ..
INTEGER, INTENT (IN) :: nobd
INTEGER, INTENT (IN) :: itype,isp,ne,lmaxb
INTEGER, INTENT (IN) :: itype,isp,ne
INTEGER, INTENT (INOUT) :: i_u ! on input: index for the first U for atom type "itype or higher"
! on exit: index for the first U for atom type "itype+1 or higher"
! ..
! .. Array Arguments ..
REAL, INTENT (IN) :: we(nobd)
COMPLEX, INTENT (IN) :: v_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u)
COMPLEX, INTENT (IN) :: v_mmp(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u)
COMPLEX, INTENT (IN) :: acof(:,0:,:)!(nobd,0:dimension%lmd,atoms%nat)
COMPLEX, INTENT (IN) :: bcof(:,0:,:)!(nobd,0:dimension%lmd,atoms%nat)
COMPLEX, INTENT (IN) :: ccof(-atoms%llod:atoms%llod,nobd,atoms%nlod,atoms%nat)
......
global/constants.f
View file @ e18fe3aa
......@@ -10,6 +10,7 @@
INTEGER, PARAMETER :: noState_const = 0
INTEGER, PARAMETER :: coreState_const = 1
INTEGER, PARAMETER :: valenceState_const = 2
INTEGER, PARAMETER :: lmaxU_const = 3
REAL, PARAMETER :: pi_const=3.1415926535897932
REAL, PARAMETER :: tpi_const=2.*3.1415926535897932
REAL, PARAMETER :: fpi_const=4.*3.1415926535897932
......
global/types.F90
View file @ e18fe3aa
......@@ -262,7 +262,7 @@ MODULE m_types
INTEGER, ALLOCATABLE :: igrd(:)
INTEGER, ALLOCATABLE :: krla(:)
LOGICAL, ALLOCATABLE :: relcor(:)
!lda_u information(ntype)
!lda_u information(size: atoms%n_u)
TYPE(t_utype),ALLOCATABLE::lda_u(:)
INTEGER,ALLOCATABLE :: relax(:,:) !<(3,ntype)
INTEGER, ALLOCATABLE :: nflip(:) !<flip magnetisation of this atom
......@@ -893,11 +893,21 @@ MODULE m_types
TYPE t_potden
! General variables and arrays
INTEGER :: iter
COMPLEX,ALLOCATABLE :: pw(:,:)
REAL,ALLOCATABLE :: mt(:,:,:,:)
REAL,ALLOCATABLE :: vacz(:,:,:)
COMPLEX,ALLOCATABLE :: vacxy(:,:,:,:)
! For density only (noco case)
COMPLEX, ALLOCATABLE :: cdom(:)
COMPLEX, ALLOCATABLE :: cdomvz(:,:)
COMPLEX, ALLOCATABLE :: cdomvxy(:,:,:)
! For density matrix and associated potential matrix
COMPLEX, ALLOCATABLE :: mmpMat(:,:,:,:)
!this type contains two init routines that should be used to allocate
!memory. You can either specify the datatypes or give the dimensions as integers
!See implementation below!
......
io/r_inpXML.F90
View file @ e18fe3aa
......@@ -119,7 +119,7 @@ SUBROUTINE r_inpXML(&
REAL :: tauTemp(3,48)
REAL :: bk(3)
LOGICAL :: flipSpin, l_eV, invSym, l_qfix, relaxX, relaxY, relaxZ
LOGICAL :: l_vca, coreConfigPresent, l_enpara, l_orbcomp
LOGICAL :: l_vca, coreConfigPresent, l_enpara, l_orbcomp, tempBool
REAL :: magMom, radius, logIncrement, qsc(3), latticeScale, dr
REAL :: aTemp, zp, rmtmax, sumWeight, ldau_u(4), ldau_j(4), tempReal
REAL :: weightScale, eParamUp, eParamDown
......@@ -1794,12 +1794,15 @@ SUBROUTINE r_inpXML(&
END IF
IF (numberNodes.EQ.1) THEN
coreSpecInput%verb = evaluateFirstBoolOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@verbose'))
coreSpecInput%verb = 0
tempBool = evaluateFirstBoolOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@verbose'))
IF(tempBool) coreSpecInput%verb = 1
coreSpecInput%ek0 = evaluateFirstOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@eKin'))
coreSpecInput%atomType = evaluateFirstIntOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@atomType'))
coreSpecInput%lx = evaluateFirstIntOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@lmax'))
coreSpecInput%edge = TRIM(ADJUSTL(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@edgeType')))
coreSpecInput%edgeidx = evaluateFirstIntOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@edgeIndex'))
coreSpecInput%edgeidx(:) = 0
coreSpecInput%edgeidx(1) = evaluateFirstIntOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@edgeIndex'))
coreSpecInput%emn = evaluateFirstOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@eMin'))
coreSpecInput%emx = evaluateFirstOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@eMax'))
tempInt = evaluateFirstIntOnly(xmlGetAttributeValue(TRIM(ADJUSTL(xPathA))//'/@numPoints'))
......
ldau/u_setup.f90
View file @ e18fe3aa
......@@ -17,12 +17,13 @@ MODULE m_usetup
! Extension to multiple U per atom type G.M. 2017 |
!-------------------------------------------------------------------+
CONTAINS
SUBROUTINE u_setup(sym,atoms,lmaxb,sphhar, input,el,vr,mpi, vs_mmp,results,number)
SUBROUTINE u_setup(sym,atoms,sphhar, input,el,pot,mpi,results)
USE m_umtx
USE m_uj2f
USE m_nmat_rot
USE m_vmmp
USE m_types
USE m_constants
IMPLICIT NONE
TYPE(t_sym),INTENT(IN) :: sym
TYPE(t_results),INTENT(INOUT) :: results
......@@ -30,14 +31,9 @@ CONTAINS
TYPE(t_input),INTENT(IN) :: input
TYPE(t_sphhar),INTENT(IN) :: sphhar
TYPE(t_atoms),INTENT(INOUT) :: atoms !n_u might be modified if no density matrix is found
TYPE(t_potden),INTENT(INOUT) :: pot
! ... Arguments ...
INTEGER, INTENT (IN) :: lmaxb
REAL, INTENT (IN) :: el(0:,:,:) !(0:atoms%lmaxd,ntype,jspd)
REAL, INTENT (IN) :: vr(:,0:,:,:) !(atoms%jmtd,0:sphhar%nlhd,ntype,jspd)
COMPLEX,INTENT (OUT)::vs_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,input%jspins)
INTEGER,INTENT(IN),OPTIONAL::number
REAL, INTENT(IN) :: el(0:,:,:) !(0:atoms%lmaxd,ntype,jspd)
! ... Local Variables ...
INTEGER itype,ispin,j,k,l,jspin,urec,i_u
INTEGER noded,nodeu,ios,lty(atoms%n_u)
......@@ -53,9 +49,6 @@ CONTAINS
! look, whether density matrix exists already:
!
filename="n_mmp_mat"
IF (PRESENT(number)) THEN
WRITE(filename,"('n_mmp_mat.',i0)") number
ENDIF
INQUIRE (file=filename,exist=n_mmp_exist)
IF (n_mmp_exist.AND.atoms%n_u>0) THEN
!
......@@ -66,21 +59,21 @@ CONTAINS
!
! set up e-e- interaction matrix
!
ALLOCATE ( u(-lmaxb:lmaxb,-lmaxb:lmaxb,&
-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,input%jspins) )
ALLOCATE (u(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,&
-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u,input%jspins))
DO ispin = 1, 1 ! input%jspins
f0(:,1) = (f0(:,1) + f0(:,input%jspins) ) / 2
f2(:,1) = (f2(:,1) + f2(:,input%jspins) ) / 2
f4(:,1) = (f4(:,1) + f4(:,input%jspins) ) / 2
f6(:,1) = (f6(:,1) + f6(:,input%jspins) ) / 2
CALL umtx(atoms,lmaxb,f0(1,ispin),&
CALL umtx(atoms,f0(1,ispin),&
f2(1,ispin),f4(1,ispin),f6(1,ispin),&
u(-lmaxb,-lmaxb,-lmaxb,-lmaxb,1,ispin) )
u(-lmaxU_const,-lmaxU_const,-lmaxU_const,-lmaxU_const,1,ispin) )
END DO
!
! read density matrix
!
ALLOCATE (ns_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,input%jspins))
ALLOCATE (ns_mmp(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u,input%jspins))
OPEN (69,file=filename,status='unknown',form='formatted')
READ (69,9000) ns_mmp
9000 FORMAT(7f20.13)
......@@ -110,8 +103,8 @@ CONTAINS
!
! calculate potential matrix and total energy correction
!
CALL v_mmp(sym,atoms,input%jspins,lmaxb,ns_mmp,u,f0,f2,&
vs_mmp,results)
CALL v_mmp(sym,atoms,input%jspins,ns_mmp,u,f0,f2,&
pot%mmpMat,results)
IF (mpi%irank.EQ.0) THEN
DO jspin = 1,input%jspins
WRITE (6,'(a7,i3)') 'spin #',jspin
......@@ -128,7 +121,7 @@ CONTAINS
ELSE
WRITE (6,*) 'using the atomic limit of LDA+U '
ENDIF
WRITE (6,l_form) ((vs_mmp(k,j,i_u,jspin),k=-l,l),j=-l,l)
WRITE (6,l_form) ((pot%mmpMat(k,j,i_u,jspin),k=-l,l),j=-l,l)
END DO
END DO
WRITE (6,*) results%e_ldau
......@@ -139,9 +132,8 @@ CONTAINS
WRITE (*,*) 'no density matrix found ... skipping LDA+U'
WRITE(*,*) "File:",filename
ENDIF
vs_mmp(:,:,:,:) = CMPLX(0.0,0.0)
pot%mmpMat(:,:,:,:) = CMPLX(0.0,0.0)
results%e_ldau = 0.0
atoms%n_u = 0
ENDIF
RETURN
......
ldau/umtx.f90
View file @ e18fe3aa
......@@ -6,7 +6,7 @@ MODULE m_umtx
!* Extension to multiple U per atom type by G.M. 2017 *
!*********************************************************************
CONTAINS
SUBROUTINE umtx(atoms,lmaxb,f0,f2,f4,f6,&
SUBROUTINE umtx(atoms,f0,f2,f4,f6,&
u)
USE m_constants
......@@ -14,16 +14,17 @@ CONTAINS
USE m_types
IMPLICIT NONE
INTEGER, PARAMETER :: lmaxw=3,lmmaxw1=(2*lmaxw+2)**2
TYPE(t_atoms), INTENT(IN) :: atoms
INTEGER, INTENT (IN) :: lmaxb
REAL, INTENT (IN) :: f0(atoms%n_u),f2(atoms%n_u),f4(atoms%n_u),f6(atoms%n_u)
REAL, INTENT (OUT) :: u(-lmaxb:lmaxb,-lmaxb:lmaxb,-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u)
TYPE(t_atoms), INTENT(IN) :: atoms
REAL, INTENT(IN) :: f0(atoms%n_u),f2(atoms%n_u),f4(atoms%n_u),f6(atoms%n_u)
REAL, INTENT(OUT) :: u(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,&
-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u)
INTEGER, PARAMETER :: lmaxw=3,lmmaxw1=(2*lmaxw+2)**2
INTEGER i,j,k,l,m,mk,nfk,itype,i_u
INTEGER m1,m2,m3,m4,lm1,lm2,lm3,lm4,kf
REAL uk,uq,avu,avj,cgk1,cgk2,tol
REAL fk(lmaxb+1,atoms%n_u)
REAL fk(lmaxU_const+1,atoms%n_u)
REAL, ALLOCATABLE :: c(:,:,:)
!
tol = 1.0e-14
......@@ -54,7 +55,7 @@ CONTAINS
END DO
END DO
CALL sgaunt(lmaxw,lmmaxw1,lmaxb,c)
CALL sgaunt(lmaxw,lmmaxw1,lmaxU_const,c)
DO i_u = 1, atoms%n_u !!! over U parameters
l = atoms%lda_u(i_u)%l
......
ldau/v_mmp.F90
View file @ e18fe3aa
......@@ -19,7 +19,7 @@ MODULE m_vmmp
! Extension to multiple U per atom type G.M. 2017
! ************************************************************
CONTAINS
SUBROUTINE v_mmp(sym,atoms,jspins,lmaxb,ns_mmp,u,f0,f2, vs_mmp,results)
SUBROUTINE v_mmp(sym,atoms,jspins,ns_mmp,u,f0,f2, vs_mmp,results)
USE m_types
USE m_constants
......@@ -29,13 +29,14 @@ CONTAINS
TYPE(t_atoms),INTENT(IN) :: atoms
!
! .. Arguments ..
INTEGER, INTENT (IN) :: lmaxb,jspins
REAL, INTENT (IN) :: u(-lmaxb:lmaxb,-lmaxb:lmaxb, -lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u)
REAL, INTENT (IN) :: f0(atoms%n_u),f2(atoms%n_u)
INTEGER, INTENT(IN) :: jspins
REAL, INTENT(IN) :: u(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,&
-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u)
REAL, INTENT(IN) :: f0(atoms%n_u),f2(atoms%n_u)
COMPLEX, INTENT(OUT) :: vs_mmp(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u,jspins)
COMPLEX :: ns_mmp(-lmaxU_const:lmaxU_const,-lmaxU_const:lmaxU_const,atoms%n_u,jspins)
COMPLEX :: ns_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,jspins)
COMPLEX,INTENT(OUT)::vs_mmp(-lmaxb:lmaxb,-lmaxb:lmaxb,atoms%n_u,jspins)
!
! .. Local Variables ..
INTEGER ispin,jspin,l ,mp,p,q,itype,m,i_u
REAL rho_tot,u_htr,j_htr,e_ee,ns_sum,spin_deg,e_dc,e_dcc
......
tests/CMakeLists.txt
View file @ e18fe3aa
......@@ -4,7 +4,7 @@ set(SerialParallelTests CuBulk CuBulkXML Fe_1l Fe_1lXML Fe-Atom CuBand CuBandXML
Fe_bct Fe_bctXML PTO PTOXML Fe_1l_SOC Fe_1l_SOCXML PTO-SOC PTO-SOCXML Fe_bct_SOC Fe_bct_SOCXML
GaAsMultiUForceXML SiFilmSlicePlotXML)
set(SerialOnlyTests Fe_bct_LO Fe_bct_LOXML Fe_fcc Fe_fccXML TiO2eels)
set(SerialOnlyTests Fe_bct_LO Fe_bct_LOXML Fe_fcc Fe_fccXML TiO2eels TiO2eelsXML)
set(Testdirs ${SerialParallelTests} ${SerialOnlyTests})
set(ParTestdirs ${SerialParallelTests})
......
tests/tests/TiO2eelsXML/files/fort.37 0 → 100644
View file @ e18fe3aa
0 1 430.098 4.3808E-03 0.0000E+00 5.3804E-03 0.0000E+00
0 1 431.098 1.1145E-03 0.0000E+00 2.2368E-03 0.0000E+00
0 1 432.098 5.4398E-07 0.0000E+00 2.4874E-06 0.0000E+00
0 1 433.098 5.7937E-15 0.0000E+00 3.4247E-14 0.0000E+00
0 1 434.098 2.8353E-27 0.0000E+00 1.4684E-26 0.0000E+00
0 1 435.098 8.8574E-14 0.0000E+00 4.1369E-13 0.0000E+00
0 1 436.098 2.1025E-05 0.0000E+00 9.0098E-05 0.0000E+00
0 1 437.098 4.6458E-02 0.0000E+00 1.4574E-01 0.0000E+00
0 1 438.098 1.9176E-01 0.0000E+00 2.2600E-01 0.0000E+00
0 1 439.098 7.0534E-02 0.0000E+00 1.1324E-01 0.0000E+00
0 1 440.098 7.8333E-02 0.0000E+00 3.2395E-01 0.0000E+00
0 1 441.098 7.8703E-02 0.0000E+00 1.5333E-01 0.0000E+00
0 1 442.098 2.8849E-02 0.0000E+00 1.2373E-01 0.0000E+00
0 1 443.098 8.9866E-04 0.0000E+00 5.0424E-03 0.0000E+00
0 1 444.098 4.6961E-03 0.0000E+00 1.9480E-02 0.0000E+00
0 1 445.098 1.7447E-03 0.0000E+00 1.9236E-03 0.0000E+00
0 1 446.098 7.4332E-03 0.0000E+00 7.6095E-03 0.0000E+00
1 1 430.098 6.3784E-02 0.0000E+00 3.5411E-02 0.0000E+00
1 1 431.098 1.4520E-02 0.0000E+00 1.4683E-02 0.0000E+00
1 1 432.098 3.7277E-06 0.0000E+00 1.6287E-05 0.0000E+00
1 1 433.098 2.0842E-14 0.0000E+00 2.2391E-13 0.0000E+00
1 1 434.098 1.7334E-26 0.0000E+00 9.6432E-26 0.0000E+00
1 1 435.098 6.6101E-13 0.0000E+00 2.7179E-12 0.0000E+00
1 1 436.098 1.7587E-04 0.0000E+00 5.9129E-04 0.0000E+00
1 1 437.098 5.1795E-01 0.0000E+00 9.5547E-01 0.0000E+00
1 1 438.098 3.0585E+00 0.0000E+00 1.4812E+00 0.0000E+00
1 1 439.098 1.0505E+00 0.0000E+00 7.4181E-01 0.0000E+00
1 1 440.098 6.7892E-01 0.0000E+00 2.1229E+00 0.0000E+00
1 1 441.098 1.1043E+00 0.0000E+00 1.0042E+00 0.0000E+00
1 1 442.098 2.3789E-01 0.0000E+00 8.1016E-01 0.0000E+00
1 1 443.098 4.3919E-03 0.0000E+00 3.2996E-02 0.0000E+00
1 1 444.098 4.0029E-02 0.0000E+00 1.2750E-01 0.0000E+00
1 1 445.098 2.6406E-02 0.0000E+00 1.2632E-02 0.0000E+00
1 1 446.098 1.1761E-01 0.0000E+00 4.9857E-02 0.0000E+00
tests/tests/TiO2eelsXML/files/inp-2.xml 0 → 100644
View file @ e18fe3aa
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<fleurInput fleurInputVersion="0.28">
<comment>
TiO2 Rutile
</comment>
<calculationSetup>
<cutoffs Kmax="3.80000000" Gmax="9.0" GmaxXC="8.00000000" numbands="40"/>
<scfLoop itmax="1" minDistance=".00000000" maxIterBroyd="99" imix="Anderson" alpha=".05000000" spinf="2.00000000"/>
<coreElectrons ctail="T" frcor="F" kcrel="0" coretail_lmax="99"/>
<magnetism jspins="1" l_noco="F" l_J="F" swsp="F" lflip="F"/>
<soc theta=".00000000" phi=".00000000" l_soc="F" spav="F" off="F"/>
<nocoParams l_ss="F" l_mperp="F" l_constr="F" l_disp="F" sso_opt="FFF" mix_b=".00000000" thetaJ=".00000000" nsh="0">
<qss>.0000000000 .0000000000 .0000000000</qss>
</nocoParams>
<expertModes gw="0" pot8="F" isec1="99" secvar="F"/>
<geometryOptimization l_f="F" xa="2.00000000" thetad="330.00000000" epsdisp=".00001000" epsforce=".00001000"/>
<bzIntegration valence