Make id_prop.json from vasprun.xml

[knc6]

For force-field training

[knc6]

Take a look:
https://colab.research.google.com/gist/knc6/5513b21f5fd83a7943509ffdf5c3608b/make_id_prop.ipynb

[GurjotDhaliwal]

Hi Kamal,

Thanks for the tutorial and the great codebase that your team has developed. I am using ALIGNN model for force field training. I have a quick question regarding the stress calculation in the above tutorial. The stress matrix is created by this code

   def voigt_6_to_full_3x3_stress(stress_vector):
    """
    Form a 3x3 stress matrix from a 6 component vector in Voigt notation, from ASE
    """
    s1, s2, s3, s4, s5, s6 = np.transpose(stress_vector)
    return np.transpose([[s1, s6, s5],
                         [s6, s2, s4],
                         [s5, s4, s3]])

Is there a reason to use 's6' in xy direction and 's5' in xz?

I checked the OUTCAR and also the Outcar function from Jarvis.io.vasp.output. Based on that, I think s5 should be in xy direction. Please let me know if that should not be the case.

For your reference below is the output of stress from the colab notebook.

 'stresses': [[-4.66994, -0.89562, -0.61014],
  [-0.89562, -4.41926, -0.32273],
  [-0.61014, -0.32273, -5.90909]],

Corresponding OUTCAR output is

  FORCE on cell =-STRESS in cart. coord.  units (eV):
  Direction    XX          YY          ZZ          XY          YZ          ZX
  --------------------------------------------------------------------------------------
  Alpha Z    73.83721    73.83721    73.83721
  Ewald    -787.95859  -785.13803  -801.96319    -3.62436    -6.91336   -10.11263
  Hartree   249.47374   248.36899   254.96190     1.41901     2.71016     3.96212
  E(xc)    -225.23462  -225.24248  -225.19559     0.01010     0.01927     0.02818
  Local    -139.75437  -141.01999  -133.47364     1.62693     3.09937     4.53620
  n-local   -84.25553   -83.96647   -85.69103    -0.37139    -0.70870    -1.03649
  augment    39.98084    39.93799    40.19366     0.05505     0.10507     0.15366
  Kinetic   873.65651   872.98165   877.00827     0.86704     1.65489     2.42008
  Fock        0.00000     0.00000     0.00000     0.00000     0.00000     0.00000
  -------------------------------------------------------------------------------------
  Total      -0.25481    -0.24113    -0.32242    -0.01761    -0.03329    -0.04887
  in kB      -4.66994    -4.41926    -5.90909    -0.32273    -0.61014    -0.89562