Descriptor

These functions are used to read initial atomic structures saved as extxyz files and create descriptors

atomdnn.descriptor.create_descriptors(elements, xyzfiles, descriptor, format='extxyz', descriptors_path=None, descriptor_filename='dump_fp.*', der_filename='dump_der.*', start_file_id=1, image_num=None, skip=0, keep_lmpfiles=False, create_data=True, verbose=False, silent=False, **kwargs)[source]

Read extxyz files as inputs and create descriptors and their derivatives w.r.t. atom coordinates.

Parameters:
  • elements – a list of elements, e.g. [‘C’,’O’,’H’], make sure the sequence is consistant

  • xyzfiles – a serials of extxyz files of input atomic structures, wildcard * is used for files numerically ordered

  • descritpor (dictionary) – the parameter dictionary of the descriptor

  • format – ‘lammp-data’,’extxyz’,’vasp’ etc. See complete list on https://wiki.fysik.dtu.dk/ase/ase/io/io.html#ase.io.read. ‘extxyz’ is recommanded.

  • descriptors_path – a new directory where descirptors will be generated, default is ‘./descriptors’

  • descriptor_filename – default is ‘dump_fp.*’, numerically ordered

  • der_filename – default is ‘dump_der.*’, numerically ordered

  • start_file_id (int) – starting id for descriptor and derivative files

  • image_num – number of images that will be used, if it’s None then read all files specified by xyzfile_name

  • skip (int) – skip some images

  • keep_lmpfiles (bool) – set to True if want to keep the lammps input and datafiles used for creating descriptors

  • create_data (bool) – set to True if want to create Data object using the generated descriptors

  • verbose (bool) – set to True if want to print out the extxyz file names used for creating descriptors

  • kwargs – used to pass optional file styles

atomdnn.descriptor.create_lmp_input(descriptor, descriptors_path)[source]

Creates a lammps input file for computing descriptors and derivatives

Parameters:

descritpor (dictionary) – the parameter dictionary of the descriptor

atomdnn.descriptor.get_filenames(file_path, file_name)[source]

Get the names of a set of files that match the patten given by file_name. Return the filenames having the format of ‘string1*string2’, in which string1 and string2 can be any strings amd * must be an integer.