Gabriel WlazłowskiGeneral info
The time-dependent code td-wslda evolves wave-functions provided from outside. For test purposes, the solver can be initialized by uniform solution, in the same way as in the case of static solver st-wslda, for more info see here.
In practical applications solution generated by st-wslda codes is used as starting point for the evolution. Figure below shows relations between initial state generators and time revolvers.
Writing wave-functions to files
In order to be able to use a solution generated by the st-wslda codes as input for td-wslda simulation user must store wave-functions in binary files. Following tags in input file control :
# Tags from st-wslda input file
outprefix st-run # all output file with start with this prefix
writewf 1 # write wf at the end of computation yes=1, no=0
iogroups 8 # number of IO groups used for parallel wf writing, default=1
# It indicates the number of processes that can write to files simultaneously.
# Note: Too many iogroups may degrade writing performance. After the run is completed in folder outprefix you will find many binary files with extensions .wfu, .wfv, .en, etc. They contain wave-functions.
See here for more info related to the processing of wave-functions.
Reading wave-functions by time-dependent codes
In order to read the wave-functions by td-wslda code you need to set following flags in the input file:
# Tags from td-wslda input file
inittype 2 # Select:
# 1 - start from st-wslda-1d solution, inprefix points to folder with s1dpca binary files
# 2 - start from st-wslda-2d solution, inprefix points to folder with s2dpca binary files
# 3 - start from st-wslda-3d solution, inprefix points to folder with s3dpca binary files
inprefix st-run # point to folder with binary files, generated by st-wslda code
# no "/" at the end
iogroups 8 # number of IO groups used for parallel reading of data
# it must be the same as the value used for static calculations.Reading process in reported in stdout, for example:
...
# ST-WSLDA-2D: file_name=`../st-testcase-uniform/test/s2dpca.info`
# ST-WSLDA-2D: nwf (with -kz's)=982
# ST-WSLDA-2D: nx=8, ny=10, nz=12, dx=1.000000, dy=1.000000, dz=1.000000
# ST-WSLDA-2D: kF=1.025823, mu_a=-0.031347, mu_b=0.489476, ec=4.907524, beta=190.057402
# ST-WSLDA-2D: nwf in binary files=552
# INIT2: nwf=552 wave-functions to scatter
# INIT2: BLOCK ID[0] CONSITING WITH 32 PROCESSES READS DATA...
# INIT2: LOADING POTENTIALS `../st-testcase-uniform/test/s2dpca.pud`...
# INIT2: TOTAL NUMBER OF PARTICLES: SPIN_A= 17 SPIN_B= 18 TOTAL= 35
...Control sums
To check the correctness of transferring wave-functions from st-wslda to td-wslda code compare content of check.stamp files. This file contains control sums, which are integrated quantities, like densities and energies, which are recalcualated from wave-functions. They shoule agree up to machine precision. For example:
[gabrielw@dell st-my-project]$ cat st-run_check.stamp
CHECK STAMP DATE: 02/05/21-09:36:32
SUM(DENSITY[ 0])= 20.333313
SUM(DENSITY[ 1])= 20.333313
SUM(DENSITY[ 2])= 17
SUM(DENSITY[ 3])= 19.139264
SUM(DENSITY[ 4])= 0
SUM(DENSITY[ 5])= 0
SUM(DENSITY[ 6])= 0
SUM(DENSITY[ 7])= 18
SUM(DENSITY[ 8])= 19.77459
SUM(DENSITY[ 9])= 0
SUM(DENSITY[10])= 0
SUM(DENSITY[11])= 0
ENERGY[ 0])= 21.27885025
ENERGY[ 1])= -5.82010539
ENERGY[ 2])= -10.64741289
ENERGY[ 3])= 0.00000000
ENERGY[ 4])= 0.00000000
ENERGY[ 5])= 0.00000000
ENERGY[ 6])= 0.00000000[gabrielw@dell td-my-project]$ cat td-run_check.stamp
CHECK STAMP DATE: 02/05/21-13:30:06
SUM(DENSITY[ 0])= 20.333313
SUM(DENSITY[ 1])= 20.333313
SUM(DENSITY[ 2])= 17
SUM(DENSITY[ 3])= 19.139264
SUM(DENSITY[ 4])= 1.5239524e-17
SUM(DENSITY[ 5])= -4.8037775e-17
SUM(DENSITY[ 6])= 0
SUM(DENSITY[ 7])= 18
SUM(DENSITY[ 8])= 19.77459
SUM(DENSITY[ 9])= 1.2655183e-17
SUM(DENSITY[10])= -2.2704154e-17
SUM(DENSITY[11])= 0
ENERGY[ 0])= 21.27885025
ENERGY[ 1])= -5.82010539
ENERGY[ 2])= -10.64741289
ENERGY[ 3])= 0.00000000
ENERGY[ 4])= 0.00000000
ENERGY[ 5])= 0.00000000
ENERGY[ 6])= 0.00000000
ENERGY[ 7])= 16.99999998
ENERGY[ 8])= 18.00000002
ENERGY[ 9])= 0.00000000
ENERGY[10])= -0.00000000