LAMMPS_methods

libra_py.LAMMPS_methods.compute_dynmat(lmp, filename, atoms, dr, opt=1)[source]

The function to return the dynamic matrix for a system using LAMMPS

Note

To use this code, one needs to have pylammps installed

Parameters

  • lmp (lammps.lammps() object) – an instance of PyLAMMPS - you need to create if first

  • filename (string) – LAMMPS input filename defining the system and the interactions (importantly! assuming everything is in ATOMIC UNITS - “units electron”). Furthermore, this file should define the minimization procedure e.g. “minimize 0.0 1.0e-8 1000 100000”

  • atoms (list of integers) – indices of the atoms for which to compute Hessian

  • dr (double) – the magnitude of the displacement [ units: Bohr ]

Returns

( D, H, R, M, E ), where:

  • D ( MATRIX(ndof, ndof) ): a sub-block of a dynamic matrix (Hessian/sqrt(m_ij))

  • H ( MATRIX(ndof, ndof) ): a sub-block of a Hessian matrix [ units: a.u.]

  • r ( MATRIX(ndof, 1) ): coordinates of all atoms [ units: Bohr ]

  • M ( MATRIX(ndof, 1) ): masses of all atoms [ units: a.u. ]

  • E ( list of ndof ints ): types of atoms - later can be mapped to atom names

  • opt ( int ): flag to use all atoms or only the pupplied sub-set

    • opt = 0: use all atoms

    • opt = 1: use only the supplied subset [ default ]

Example

>> import lammps >> lmp = lammps.lammps() >> D, H, R, M, E = compute_dynmat(lmp, “in.lammps”, [0,1,2], 1e-3) >> print “Atom types are:”, E >> print “Masses are:”; M.show_matrix() >> print “Coordinates are:”; R.show_matrix() >> print “Hessian matrix is:”; H.show_matrix() >> print “Dynamical matrix is:”; D.show_matrix()

Return type

tuple