This module deals with direct data exchange between the LAMMPS simulation code and OVITO. The following function allows employing OVITO’s data analysis and rendering functions within a running molecular dynamics simulation.


The function from this module is typically called from a Python script being executed in the context of a running LAMMPS simulation. Please install the ovito package in the Python interpreter used by your LAMMPS installation to make it available for import.

ovito.io.lammps.lammps_to_ovito(lmp, data_collection=None)

Constructs an OVITO DataCollection from a LAMMPS simulation state. This function should be called from within a running LAMMPS molecular dynamics simulation to hand over the current particle model to OVITO and perform data analysis or visualization tasks in regular time intervals, see the usage example below.

  • lmp – The LAMMPS Python object, which is used to by the function to access the particle data of the simulation.

  • data_collection – An existing DataCollection to fill in with the particle model. A new data collection will be created if not provided.


DataCollection containing the converted particle data - only on rank 0 of a parallel LAMMPS simulation. On other ranks, the function returns None.


This function is still under active development and should be considered as experimental. Only the simulation cell geometry, boundary conditions, and particle coordinates are converted from LAMMPS to the OVITO representation. Please contact the OVITO developers if you require more particle properties to be transferred from LAMMPS for your specific application.

Usage example:

In the LAMMPS simulation script, include the following python and fix python/invoke commands to load the following Python script into the context of the LAMMPS simulation and invoke the perform_dxa() function in regular timestep intervals during the MD run:

python perform_dxa source perform_dxa.py
fix 1 all python/invoke 1000 end_of_step perform_dxa

The following script file, perform_dxa.py, defines the Python function to be executed every 1000 timesteps. It runs OVITO’s DislocationAnalysisModifier on the current simulation snapshot to identify all dislocation defects in the crystal:

from lammps import lammps
from ovito.io.lammps import lammps_to_ovito
from ovito.io import export_file
from ovito.modifiers import DislocationAnalysisModifier

def perform_dxa(lammps_ptr):

    # Access the LAMMPS simulation state.
    lmp = lammps(ptr=lammps_ptr)

    # Convert simulation state to OVITO data representation.
    data = lammps_to_ovito(lmp)

    # Let OVITO perform data analysis only on MPI rank 0. Do nothing on other processors.
    if data is None: 

    # Run DXA.

    # Output total length of all dislocation lines.
    total_line_length = data.attributes['DislocationAnalysis.total_line_length']
    print('Total dislocation line length:', total_line_length)

    # Dump dislocation lines to a CA file for later visualization in the OVITO desktop application.
    timestep = int(lmp.get_thermo('step'))
    export_file(data, f'dislocations.{timestep}.ca', format='ca')