Question about the “map_particles_to_regions” issue in ConstructSurfaceModifier

OVITO
1

Dear forum members,

I have encountered a problem while using the ConstructSurfaceModifier function in my code. Specifically, I am trying to construct a surface using certain particles from the file “dump_initial_config.colloid”(see below) using the alpha shape method. The IDs of these particles are listed in the file “special_cluster_atoms”. Although these particles do not form a compact shape, they do form a cluster.
special_cluster_atoms

To construct the surface using these particles, I have used the option “only_selected=True” in the ConstructSurfaceModifier function. Additionally, I want to check if other particles are inside the surface by using the option “map_particles_to_regions=True” in the same function.

Following the example in the OVITO documentation, I attempt to count the number of particles inside the surface. However, I encounter the following error:

in_surface = np.count_nonzero(surface_region[‘Filled’][data.particles[“Region”]])
IndexError: index 5 is out of bounds for axis 0 with size 5

Below is the relevant portion of my code:

import ovito
import numpy as np

def main():
    special_atoms_id = np.loadtxt(“./special_cluster_atoms”, skiprows=9)
    dump_file = “./dump_initial_config.colloid”

    pipeline = ovito.io.import_file(dump_file)
    pipeline.modifiers.append(ovito.modifiers.ExpressionSelectionModifier(expression="%s"convert_ovito_command(special_atoms_id)))
    pipeline.modifiers.append(
    ovito.modifiers.ConstructSurfaceModifier(method=ovito.modifiers.ConstructSurfaceModifier.Method.AlphaShape,
                                             radius=1.6,
                                             identify_regions=True,
                                             map_particles_to_regions=True,
                                             only_selected=True))

    data = pipeline.compute(0)
    surface_region = data.surfaces['surface'].regions
    in_surface = np.count_nonzero(surface_region['Filled'][data.particles["Region"]])

def convert_ovito_command(atoms_id_list):
    head = ‘ParticleIdentifier ==’
    joint = ‘||’
    command = ‘’
    for atom_id in atoms_id_list:
    command = command + head + str(atom_id) + joint
    command = command[:-2]
    return command

if __name__ == ‘__main__’:
    main()

Interestingly, when I increase the radius in the alpha shape method (e.g., changing it from 1.6 to 3), the error disappears. Could someone kindly provide possible solutions to this issue?

By the way, if I have a list of particle IDs that I want to select, I currently use the convert_ovito_command function to generate an expression. Is there a more convenient method to achieve this?

Thank you in advance for your assistance.

dump_initial_config.colloid (159.8 KB)
special_cluster_atoms (385 Bytes)

2

Thank you very much for your detailed error report. We have been able to reproduce the error and will fix it in the next OVITO release.

3

Concerning your second question, this is an alternative way of selecting particles according to the list of particle identifiers:

    ids = np.loadtxt(...)
    sel = data.particles_.create_property("Selection")
    sel[np.isin(data.particles.identifiers, ids)] = 1

which you could do within the scope of a python script modifier that is part of the pipeline.

4

Thank you for your reply. May I know when the next version of Python-OVITO is expected to be released? Is the notification posted on the official website?

5

We haven’t set a release date for OVITO 3.10 yet. It’ll be later this year. But we can make a pre-release version of the Python package available to you after the issue has been fixed. Do you use the pip or the conda version of the Python package?

6

I installed Python-OVITO using pip and the current version is 3.9.2. Python version is 3.10.5.

7

Updated development builds of the Python package are available now here:

8

Thank you all very much for your help.