INTERACTIVE: Difference between revisions

From VASP Wiki
No edit summary
No edit summary
Line 7: Line 7:


  vasp_std < POSCAR.interactive
  vasp_std < POSCAR.interactive
 
{{NB|important|A {{FILE|POSCAR}} file is still required for the lattice. The calculation will not run without it.}}
{{NB|important|ADD POSCAR}}


The input structure (e.g., <code>POSCAR.interactive</code>, or any other name) is as follows:
The input structure (e.g., <code>POSCAR.interactive</code>, or any other name) is as follows:

Revision as of 12:20, 11 February 2026

INTERACTIVE = [logical]
Default: INTERACTIVE = .FALSE. 

Description: Select the use of interactive mode.


This flag determines whether or not interactive mode should be used. Interactive mode is executed by inputting a series of ionic coordinates into the VASP executable, i.e.:

vasp_std < POSCAR.interactive
Important: A POSCAR file is still required for the lattice. The calculation will not run without it.

The input structure (e.g., POSCAR.interactive, or any other name) is as follows:

  0.51602654  0.60200207  0.48355839
  0.47803882  0.52340268  0.50869036
  0.56717477  0.65578242  0.53100206
  0.45116332  0.63676166  0.43537938
  0.31530340  0.74388198  0.64715720
  0.60071504  0.49851047  0.37872126

  0.44216661  0.56361173  0.52960446
  0.36537533  0.54238027  0.56342416
  0.50398907  0.58877046  0.59064245
  0.43618126  0.61788131  0.46024981
  0.45532341  0.84599587  0.53226938
  0.50724841  0.41695239  0.46229896

  0.53802286  0.56353392  0.51036499
  0.47205503  0.63101620  0.50503092
  0.55908887  0.54004979  0.59586980
  0.61484211  0.57816646  0.45750405
  0.42364771  0.83966876  0.53596644
  0.46803897  0.42328326  0.47142822

with the coordinates of the ions for each structure given, followed by a blank line, then the next structure, etc. These calculations will then perform on these structures. The number of ionic steps NSW should be set to the number of structures in the POSCAR.interactive file.

Tip

  • We suggest using it to systematically improve machine-learned force fields (MLFF) by selecting the structures for which the MLFF has broken down and continuing to train the MLFF with them. These structures are those where the spilling factor (cf. ML_ESTBLOCK) deviates from 0, approaching 1.

Related tags and articles