QuantumATK Forum
QuantumATK => Scripts, Tutorials and Applications => Topic started by: premkumar85 on July 1, 2009, 11:06
-
I have downloaded a tutorial from quantumwise website about transport properties of gold mono wires. I made the two probe with mono wire geometry and saved it as python file. first i did self-consistent calculation and saved that checkpoint file as 'gold.nc' in the same directory in which i have all my scripts. then i started to run the script to calculate the necessary k-points and got error message "NLValueError: The self-consistent calculation must contain a two-probe system"
I have attached that k-point calculation script with this message
you can see that tutorial in this link http://www.quantumwise.com/documents/tutorials/Tutorial_GoldMonowire.pdf
please help me to solve this error
-
This error can only mean one thing, that the checkpoint (NC) file you are referring to is not a two-probe calculation (or, that the two-probe calculation did not converge).
Perhaps you have mixed up two different file? If possible, please attach the log and input script of the two-probe calculation, perhaps there is some minor typo or so...
-
i have attached the scripts for self consistent calculation and k-point calculation. i also attached self consistent calculation output vnl file.
i used Result Browser to view scripts and vnl files so i copied them into note pad documents
How do we recognize that the system is not yet converged?
-
here is the log for k-point caculation
-
Any chance you have the log for the 2probe scf calculation? That's where we could see if it converged or not.
PS: If possible, please keep the extension of the files, like .py for scripts and .vnl for VNL files. Makes it easier to download them :)
-
this is the log for self consistent calculation
-
This calculation has not finished. Either it's still running, or it was terminated (perhaps due to out of memory, it's quite a big system).
-
this is the log for self consistent calculation
As you may know, a converged self-consistent calculation of two-probe system usually contain three stages according to the sequence: i) the self-consistent calculation for electrode; ii) the self-consistent calculation for an equivalent bulk system; iii) the self-consistent calculation for the two-problem system.
Your log file for self-consistent calculation indicates your job is still on the stage of "Equivalent Bulk Calculation (Initial Density for TwoProbe)", so the corresponding 'gold.nc' file just contains the results of the equivalent bulk system and other information for the two electrodes. This 'gold.nc' can not be restored for the calculation of transmission spectrum.
To recognize the convergence of self-consistent calculation of a two-probe system, it is easy to do by first checking which stage the calculation stops at, and then by tracking the convergence criterion (e.g. the charge density change in each iteration step. the actual physical quantity depends on what you have set in the input file to control the convergence of self-consistent calculation.).