Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Anders Blom

Pages: 1 2 3 [4] 5 6 ... 361
46
No, your images were still correctly interpolated with IDPP in the old version too. It's just that in the new version we provide more options to re-interpolate the path, and it is then convenient to script the keywords needed, as a template.

Now, there is a small practical consequence to doing that, because actually it will always do a re-interpolation if the keyword generate_images is not False, and moreover it does a small random perturbation when doing that. So actually, in the new version, the initial path you will be running with is slightly different from the one in the script. The perturbation is however not random if you run multiple times, as the random seed is fixed. We might improve this in future versions, to make the re-interpolation more explicit on-demand.

47
General Questions and Answers / Re: Settings LJ for Tags
« on: January 12, 2024, 21:17 »
You still need to have pairwise interaction between atoms to define the potential, but tags can be used to define which atoms interact (so that L and L atoms don't interact).

48
General Questions and Answers / Re: Phonon spectrum error
« on: January 12, 2024, 21:11 »
What error? In the GUI?

49
General Questions and Answers / Re: Questions about nlread
« on: January 12, 2024, 21:05 »
Yeah, the second argument to nlread should not be the object id but the class type, so instead of

Device_configuration = nlread('Si80_111_5e19_4040.hdf5',DeviceConfiguration_0)[0]

do

Device_configuration = nlread('Si80_111_5e19_4040.hdf5',DeviceConfiguration)[0]

Sorry, I didn't notice this mistake earlier.

50
General Questions and Answers / Re: Optimize Cell Error
« on: January 12, 2024, 21:03 »
That makes no sense, since this keywords is implemented in that version. Can you share the full script, and the full output?

51
I do agree with AsifShah though, that it seems to be of limited usefulness, but it can be done (not sure about "automatically" but you can use a hook function).

You basically just need a post_step_hook function and pass this as argument in OptimizeGeometry:

Code: python
def computeMulliken(step, config):
    m = MullikenPopulation(config)
    nlsave("some_file.hdf5", m)

opt_geo = OptimizeGeometry(
    configuration=configuration,
    max_forces=0.01*eV/Angstrom,
    ...
    post_step_hook = computeMulliken,
    ....
)

52
No, we haven't been motivated to implement this as it doesn't seem particularly accurate.

53
You can read more about it here: https://docs.quantumatk.com/tutorials/dft_half_pps/dft_half_pps.html. In short, it's an attempt to bring some aspect of empirical pseudopotentials into DFT, but instead of a scissor operator acting on the final band structure (which wouldn't work in Ge for instance) there are shifts being applied within the self-consistent cycle. It does require empirical tuning, i.e. you should know the band gap from experiments to tune the parameters.

We implemented this method quite a long time ago, and honestly have not found much use for it. Since then, we have added DFT+1/2 which seems to be a more robust and general scheme, where the same parameters seem to work for many different systems, although you can fine tune them if you do know the band gap. And of course HSE06 and other hybrid functionals which perform really well in LCAO, and which can be truly predictive of the band gaps of unknown materials.

54
I'm afraid we only have MullikenPopulations, which obviously isn't as advanced. But it depends a bit on the end goal - you could perhaps use ProjectedDensityOfStates if you are looking to see where different orbitals or sites are positioned in energy

55
General Questions and Answers / Re: Questions about nlread
« on: January 5, 2024, 20:43 »
The code is a bit confusing, it shouldn't run actually.

I am assuming that the file Si80_111_5e19_4040.hdf5 contains one DeviceConfiguration, and that is is converged. If so, you do not need to set a calculator on it, and run update. In fact, that will re-run it (at least if any of the calculator setting are different). But in your code, the variable "calculator" is not defined, so that would give an error.

Second, you are already selecting the first object in the list in the line
Device_configuration = nlread('Si80_111_5e19_4040.hdf5',DeviceConfiguration_0)[0]
so the line
Device_configuration = Device_configuration_list[0]
is not needed, and also will give an error since Device_configuration_list is not defined.

And there is no need to have a second nlsave right after you just read the file...

So, all you really need is
Device_configuration = nlread('Si80_111_5e19_4040.hdf5',DeviceConfiguration_0)[0]
and then go straight into the part under
# %% TransmissionSpectrum

56
No, this is not implemented in QuantumATK.

57
You can also use ProjectedDensityOfStates (https://docs.quantumatk.com/manual/Types/ProjectedDensityOfStates/ProjectedDensityOfStates.html) for this, but in both cases we are referring to running a QuantumATK calculation. Planewave DFT like VASP can't make projections on angular momenta as easily as we do in LCAO-DFT, but you can ask for projections in DOSCAR in VASP, and it will show up in the analyzer in NanoLab.

58
General Questions and Answers / Re: Error: TIME LIMIT
« on: January 5, 2024, 20:33 »
This is related to your local cluster setup, not QuantumATK. Please ask your sysadmin

59
My guess is that you might run out of memory, it seems like a large calculation (many atoms). But it's very hard to tell without details.

60
General Questions and Answers / Re: Optimize Cell Error
« on: January 5, 2024, 20:07 »
The optimize_cell keyword was implemented in 2022.03 so perhaps you are by mistake running the script with an older version of atkpython, even if the script is generated with a newer GUI version?

Pages: 1 2 3 [4] 5 6 ... 361