[SOLVED] line minimization not converged after 13 restarts
Posted: Mon Feb 22, 2010 3:22 pm
Hello,
this error used to appear frequently because of the conjugate gradient method used to get the ground state wavefunctions (iscf 5). It means that the CG solver is stuck, and a quadratic fit to the energy cannot find a reasonable prediction for the next SCF step (after 13 tries) and it stops.
This has been superseded by iscf 7 (the default) which never gets into this state. Use that instead. For PAW potentials use the default iscf 17 (basically you can always leave out iscf - the default is the best).
Matthieu
this error used to appear frequently because of the conjugate gradient method used to get the ground state wavefunctions (iscf 5). It means that the CG solver is stuck, and a quadratic fit to the energy cannot find a reasonable prediction for the next SCF step (after 13 tries) and it stops.
This has been superseded by iscf 7 (the default) which never gets into this state. Use that instead. For PAW potentials use the default iscf 17 (basically you can always leave out iscf - the default is the best).
Matthieu
hello
I ran rf1 lesson for LaCoO3
my input variables are:
# Crystalline LaCoO3(cubic) : computation of the total energy
#
#Specific to ground state calculation
kptopt 1
tolvrs 1.0d-18
iscf 5
#######################################################################
#Common input variables
#Definition of the unit cell
acell 3*7.22
rprim 1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
#Definition of the atom types
ntypat 3
znucl 57 27 8
#Definition of the atoms
natom 5
typat 1 2 3 3 3
xred
0.0 0.0 0.0
0.5 0.5 0.5
0.5 0.5 0.0
0.5 0.0 0.5
0.0 0.5 0.5
#Gives the number of band, explicitely (do not take the default)
nband 15
#Exchange-correlation functional
ixc 1
#Definition of the planewave basis set
ecut 55.0
#Definition of the k-point grid
kptrlatt -4 4 4
4 -4 4
4 4 -4
#Definition of the SCF procedure
nstep 55
diemac 9.0
and the error in log file is:
,Min el dens= 7.6096E-05 el/bohr^3 at reduced coord. 0.0000 0.0000 0.0000
,Max el dens= 1.3258E+00 el/bohr^3 at reduced coord. 0.5600 0.5600 0.5600
rhohxc_coll : enter with option, nspden 1 1
ETOT 46 -85.312003083360 -8.606E-05 1.641E-04 1.992E-01 0.000E+00 0.000E+00
scprqt: <Vxc>= -4.3537831E-01 hartree
findmin : lambda_predict -7.9456E-02 etotal_predict -8.5312079788E+01
scfcge : ERROR -
Potential-based CG line minimization not converged after 13 restarts.
Action : read the eventual warnings about lack of convergence.
Some might be relevant. Otherwise, raise nband.
leave_new : decision taken to exit ...
1
and if I increase the nband, in 3rd program in rf1 lesson an error occurs because of this increasing.
how can I solve this problem?
thanks