Source code for pina.geometry.intersection_domain

"""Module for Intersection class. """

import torch
from ..label_tensor import LabelTensor
from .operation_interface import OperationInterface
import random




class Intersection(OperationInterface):

    def __init__(self, geometries):
        r"""
        PINA implementation of Intersection of Domains.
        Given two sets :math:`A` and :math:`B` then the
        domain difference is defined as:

        .. math::
            A \cap B = \{x \mid x \in A \land x \in B\},

        with :math:`x` a point in :math:`\mathbb{R}^N` and :math:`N`
        the dimension of the geometry space.

        :param list geometries: A list of geometries from ``pina.geometry``
            such as ``EllipsoidDomain`` or ``CartesianDomain``. The intersection
            will be taken between all the geometries in the list. The resulting
            geometry will be the intersection of all the geometries in the list.

        :Example:
            >>> # Create two ellipsoid domains
            >>> ellipsoid1 = EllipsoidDomain({'x': [-1, 1], 'y': [-1, 1]})
            >>> ellipsoid2 = EllipsoidDomain({'x': [0, 2], 'y': [0, 2]})
            >>> # Create a Intersection of the ellipsoid domains
            >>> intersection = Intersection([ellipsoid1, ellipsoid2])
        """
        super().__init__(geometries)



    def is_inside(self, point, check_border=False):
        """
        Check if a point is inside the ``Intersection`` domain.

        :param point: Point to be checked.
        :type point: torch.Tensor
        :param bool check_border: If ``True``, the border is considered inside.
        :return: ``True`` if the point is inside the Intersection domain, ``False`` otherwise.
        :rtype: bool
        """
        flag = 0
        for geometry in self.geometries:
            if geometry.is_inside(point, check_border):
                flag += 1
        return flag == len(self.geometries)




    def sample(self, n, mode="random", variables="all"):
        """
        Sample routine for ``Intersection`` domain.

        :param int n: Number of points to sample in the shape.
        :param str mode: Mode for sampling, defaults to ``random``. Available modes include: ``random``.
        :param variables: Variables to be sampled, defaults to ``all``.
        :type variables: str | list[str]
        :return: Returns ``LabelTensor`` of n sampled points.
        :rtype: LabelTensor

        :Example:
            >>> # Create two Cartesian domains
            >>> cartesian1 = CartesianDomain({'x': [0, 2], 'y': [0, 2]})
            >>> cartesian2 = CartesianDomain({'x': [1, 3], 'y': [1, 3]})
            >>> # Create a Intersection of the ellipsoid domains
            >>> intersection = Intersection([cartesian1, cartesian2])
            >>> # Sample
            >>> intersection.sample(n=5)
                LabelTensor([[1.7697, 1.8654],
                            [1.2841, 1.1208],
                            [1.7289, 1.9843],
                            [1.3332, 1.2448],
                            [1.9902, 1.4458]])
            >>> len(intersection.sample(n=5)
                5

        """
        if mode != "random":
            raise NotImplementedError(
                f"{mode} is not a valid mode for sampling."
            )

        sampled = []

        # calculate the number of points to sample for each geometry and the remainder.
        remainder = n % len(self.geometries)
        num_points = n // len(self.geometries)

        # sample the points
        # NB. geometries as shuffled since if we sample
        # multiple times just one point, we would end
        # up sampling only from the first geometry.
        iter_ = random.sample(self.geometries, len(self.geometries))
        for i, geometry in enumerate(iter_):
            sampled_points = []
            # int(i < remainder) is one only if we have a remainder
            # different than zero. Notice that len(geometries) is
            # always smaller than remaider.
            # makes sure point is uniquely inside 1 shape.
            while len(sampled_points) < (num_points + int(i < remainder)):
                sample = geometry.sample(1, mode, variables)
                if self.is_inside(sample):
                    sampled_points.append(sample)
            sampled += sampled_points

        return LabelTensor(torch.cat(sampled), labels=self.variables)