Dear Nordland,
I tried to run again but the out put *nc file still cannot open...
if the parallel system is working well it means that the problem may be come from the input file....
Due that I will explain clearly all steps I did ... could you please check for me ... ...
My situation is: using VNL on separated machine.
- making input file (*Py file) on PC -windows xp 64 bit from my lab
- then using Cmd file to upload input file on Parallel system -linux based 64 bit in other place.
1- Making input file:
using ATK.12.2.1 to build the py file, and then change the "nlsave" which looks like this:
------
--------------
#----------------------------------------
# Device Calculator
#----------------------------------------
calculator = DeviceLCAOCalculator(
basis_set=basis_set,
numerical_accuracy_parameters=device_numerical_accuracy_parameters,
electrode_calculators=
[left_electrode_calculator, right_electrode_calculator],
electrode_voltages=( 0.6*Volt, 0.0*Volt)
)
device_configuration.setCalculator(calculator)
nlprint(device_configuration)
device_configuration.update()
nlsave('/pwork01/p037jhd/Hien/test02/test.nc', device_configuration)
# -------------------------------------------------------------
# Transmission spectrum
# -------------------------------------------------------------
transmission_spectrum = TransmissionSpectrum(
configuration=device_configuration,
energies=numpy.linspace(-3,3,101)*eV,
kpoints=MonkhorstPackGrid(1,1),
energy_zero_parameter=AverageFermiLevel,
infinitesimal=1e-06*eV,
self_energy_calculator=DirectSelfEnergy(),
)
nlsave('/pwork01/p037jhd/Hien/test02/test.nc', transmission_spectrum, labels=['TS'])
nlprint(transmission_spectrum)
# -------------------------------------------------------------
# Electron difference density
# -------------------------------------------------------------
electron_difference_density = ElectronDifferenceDensity(device_configuration)
nlsave('/pwork01/p037jhd/Hien/test02/test.nc', electron_difference_density, labels=['EDD'])
---------
-------------
2- upload file on paralell system by Cmd file:
-----
-------
#!/bin/bash
#@ job_type = MPICH
#@ wall_clock_limit = 120:00:00
#@ step_name = test.py
#@ notification = complete
#@ class = normal
#@ error = $(step_name).$(jobid).$(stepid).err
#@ output = $(step_name).$(jobid).$(stepid).out
#@ resources = ConsumableCpus(1)
##@ total_tasks = 6
##@ blocking = unlimited
#@ node = 6
#@ tasks_per_node = 16
#@ queue
. /home/applic/compilers/intel/2012.0.032/composer_xe_2011_sp1.6.233/bin/compilervars_arch.sh intel64
. /home/applic/compilers/intel/2012.0.032/impi/4.0.3.008/bin64/mpivars.sh intel64
. /home/applic/compilers/intel/2012.0.032/mkl/bin/mklvars.sh intel64
time mpirun -np $LOADL_TOTAL_TASKS -machinefile $LOADL_HOSTFILE /home/applic/Applications/QuantumWise/atk-12.2.1/atkpython/bin/atkpython
/pwork01/p037jhd/Hien/test02/test.py--------------
--------
then submit cmd file on parallel system.
3- Calculation:
after calculation I got the *out file looks like the log file, which shows:
----
----
+----------------------------------------------------------+
Timing: Total Per Step %
--------------------------------------------------------------------------------
Density Matrix : 458.90 s 14.34 s 33.78% |=============|
Setting Density Matrix : 199.01 s 49.75 s 14.65% |=====|
Real Space Integral : 166.66 s 2.11 s 12.27% |====|
Valence Density : 118.98 s 1.51 s 8.76% |===|
Transmission Spectrum : 67.90 s 67.90 s 5.00% |=|
Diagonalization : 60.92 s 1.42 s 4.48% |=|
Difference Density : 42.31 s 0.92 s 3.11% ||
Core Density : 37.15 s 0.81 s 2.73% ||
Hartree Potential : 34.32 s 0.43 s 2.53% ||
Exchange-Correlation : 21.31 s 0.27 s 1.57% ||
Constant Terms : 20.31 s 10.15 s 1.49% ||
Mixing : 15.54 s 0.21 s 1.14% |
Basis Set Generation : 12.68 s 3.17 s 0.93% |
Real Space Basis : 7.71 s 1.93 s 0.57% |
Neutral Atom Potential : 2.73 s 0.91 s 0.20% |
File IO, nlsave : 0.20 s 0.07 s 0.01% |
Hubbard Term : 0.00 s 0.00 s 0.00% |
--------------------------------------------------------------------------------
Total : 1358.68 s (22m38.68s)
----------
-------------
so I think the calculation was done!
automatically I obtained the NC file (around 20mb) .... and I downloaded and saved in my PC...
reopen VnL windows and run analysis file and that shows the error.
.....
so what should I do?
can you hepl me?
Show Posts
