Source code for freud.data

# Copyright (c) 2010-2020 The Regents of the University of Michigan
# This file is from the freud project, released under the BSD 3-Clause License.

R"""
The :class:`freud.data` module provides certain sample data sets and utility
functions that are useful for testing and examples.

.. rubric:: Stability

:mod:`freud.data` is **unstable**. When upgrading from version 2.x to 2.y (y >
x), existing freud scripts may need to be updated. The API will be finalized in
a future release.
"""

import numpy as np
import freud


[docs]class UnitCell(object): """Class to represent the unit cell of a crystal structure. This class represents the unit cell of a crystal structure, which is defined by a lattice and a basis. It provides the basic attributes of the unit cell as well as enabling the generation of systems of points (optionally with some noise) from the unit cell. Args: box: A box-like object (see :meth:`~freud.box.Box.from_box`) containing the lattice vectors of the unit cell. basis_positions ((:math:`N_{points}`, 3) :class:`numpy.ndarray`): The basis of the unit cell in fractional coordinates (Default value = ``[[0, 0, 0]]``). """ def __init__(self, box, basis_positions=[[0, 0, 0]]): self._box = freud.box.Box.from_box(box) self._basis_positions = basis_positions
[docs] def generate_system(self, num_replicas=1, scale=1, sigma_noise=0, seed=None): """Generate a system from the unit cell. The box and the positions are expanded by ``scale``, and then Gaussian noise with standard deviation ``sigma_noise`` is added to the positions. All points are wrapped back into the box before being returned. Args: num_replicas (:class:`tuple` or int): If provided as a single number, the number of replicas in all dimensions. If a tuple, the number of times replicated in each dimension. Must be of the form ``(nx, ny, 1)`` for 2D boxes (Default value = 1). scale (float): Factor by which to scale the unit cell (Default value = 1). sigma_noise (float): The standard deviation of the normal distribution used to add noise to the positions in the system (Default value = 0). seed (int): If provided, used to seed the random noise generation. Not used unless ``sigma_noise`` > 0 (Default value = :code:`None`). Returns: tuple (:class:`freud.box.Box`, :class:`np.ndarray`): A system-like object (see :class:`~freud.locality.NeighborQuery.from_system`). """ try: nx, ny, nz = num_replicas except TypeError: nx = ny = num_replicas nz = 1 if self.box.is2D else num_replicas if not all((int(n) == n and n > 0 for n in (nx, ny, nz))): raise ValueError("The number of replicas must be a positive " "integer in each dimension.") if self.box.is2D and nz != 1: raise ValueError("The number of replicas in z must be 1 for a " "2D unit cell.") if any([n > 1 for n in (nx, ny, nz)]): pbuff = freud.locality.PeriodicBuffer() abs_positions = self.box.make_absolute(self.basis_positions) pbuff.compute((self.box, abs_positions), buffer=(nx-1, ny-1, nz-1), images=True) box = pbuff.buffer_box*scale positions = np.concatenate((abs_positions, pbuff.buffer_points)) else: box = self.box*scale positions = self.box.make_absolute(self.basis_positions) # Even numbers of repeats shift the box by L/2 shift_vec = (np.array([nx, ny, nz]) + 1) % 2 positions += shift_vec * self.box.make_absolute([1, 1, 1]) positions *= scale positions = box.wrap(positions) if sigma_noise != 0: rs = np.random.RandomState(seed) mean = [0]*3 var = sigma_noise*sigma_noise cov = np.diag([var, var, var if self.dimensions == 3 else 0]) positions += rs.multivariate_normal( mean, cov, size=positions.shape[:-1]) positions = box.wrap(positions) return box, positions
@property def box(self): """:class:`freud.box.Box`: The box instance containing the lattice vectors.""" return self._box @property def lattice_vectors(self): """:math:`(3, 3)` :class:`np.ndarray`: The matrix of lattice vectors.""" return self.box.to_matrix() @property def basis_positions(self): """:math:`(N_{points}, 3)` :class:`np.ndarray`: The basis positions.""" return self._basis_positions @property def a1(self): """:math:`(3, )` :class:`np.ndarray`: The first lattice vector.""" return self.box.to_matrix()[:, 0] @property def a2(self): """:math:`(3, )` :class:`np.ndarray`: The second lattice vector.""" return self.box.to_matrix()[:, 1] @property def a3(self): """:math:`(3, )` :class:`np.ndarray`: The third lattice vector.""" return self.box.to_matrix()[:, 2] @property def dimensions(self): """int: The dimensionality of the unit cell.""" return self.box.dimensions
[docs] @classmethod def fcc(cls): """Create a face-centered cubic (fcc) unit cell. Returns: :class:`~.UnitCell`: A face-centered cubic unit cell. """ fractions = np.array([[.5, .5, 0], [.5, 0, .5], [0, .5, .5], [0, 0, 0]]) return cls([1, 1, 1], fractions)
[docs] @classmethod def bcc(cls): """Create a body-centered cubic (bcc) unit cell. Returns: :class:`~.UnitCell`: A body-centered cubic unit cell. """ fractions = np.array([[.5, .5, .5], [0, 0, 0]]) return cls([1, 1, 1], fractions)
[docs] @classmethod def sc(cls): """Create a simple cubic (sc) unit cell. Returns: :class:`~.UnitCell`: A simple cubic unit cell. """ fractions = np.array([[0, 0, 0]]) return cls([1, 1, 1], fractions)
[docs] @classmethod def square(cls): """Create a square unit cell. Returns: :class:`~.UnitCell`: A square unit cell. """ fractions = np.array([[0, 0, 0]]) return cls([1, 1], fractions)
[docs] @classmethod def hex(cls): """Create a hexagonal unit cell. Returns: :class:`~.UnitCell`: A hexagonal unit cell. """ fractions = np.array([[0, 0, 0], [0.5, 0.5, 0]]) return cls([1, np.sqrt(3)], fractions)
[docs]def make_random_system(box_size, num_points, is2D=False, seed=None): R"""Helper function to make random points with a cubic or square box. Args: box_size (float): Size of box. num_points (int): Number of points. is2D (bool): If true, creates a 2D system. (Default value = :code:`False`). seed (int): Random seed to use. (Default value = :code:`None`). Returns: tuple (:class:`freud.box.Box`, :math:`\left(num\_points, 3\right)` :class:`numpy.ndarray`): Generated box and points. """ # noqa: E501 rs = np.random.RandomState(seed) fractional_coords = rs.random_sample((num_points, 3)) if is2D: box = freud.box.Box.square(box_size) fractional_coords[:, 2] = 0 else: box = freud.box.Box.cube(box_size) points = box.make_absolute(fractional_coords) return box, points