Net/grid file

The computational grid for D-Flow FM is stored in the net file. It is an unstructured grid, where 1D and 2D can be combined. The file format is NetCDF, using the CF-, UGRID- and Deltares-conventions.

The net file is represented by the classes below.

Model

Branch

Source code in hydrolib/core/dflowfm/net/models.py

class Branch:
    def __init__(
        self,
        geometry: np.ndarray,
        branch_offsets: np.ndarray = None,
        mask: np.ndarray = None,
    ) -> None:
        # Check that the array has two collumns (x and y)
        assert geometry.shape[1] == 2

        # Split in x and y
        self.geometry = geometry
        self._x_coordinates = geometry[:, 0]
        self._y_coordinates = geometry[:, 1]

        # Calculate distance of coordinates along line
        segment_distances = np.hypot(
            np.diff(self._x_coordinates), np.diff(self._y_coordinates)
        )
        self._distance = np.concatenate([[0], np.cumsum(segment_distances)])
        self.length = segment_distances.sum()

        # Check if mask and branch offsets (if both given) have same shape
        if (
            mask is not None
            and branch_offsets is not None
            and branch_offsets.shape != mask.shape
        ):
            raise ValueError("Mask and branch offset have different shape.")

        # Set branch offsets
        self.branch_offsets = branch_offsets
        # Calculate node positions
        if branch_offsets is not None:
            self.node_xy = self.interpolate(branch_offsets)

        # Set which of the nodes are present
        if (mask is None) and (branch_offsets is not None):
            self.mask = np.full(branch_offsets.shape, False)
        else:
            self.mask = mask

    def generate_nodes(
        self,
        mesh1d_edge_length: float,
        structure_chainage: Optional[List[float]] = None,
        max_dist_to_struc: Optional[float] = None,
    ):
        """Generate the branch offsets and the nodes.

        Args:
            mesh1d_edge_length (float): The edge length of the 1d mesh.
            structure_chainage (Optional[List[float]], optional): A list with the structure chainages. If not specified, calculation will not take it into account. Defaults to None.
            max_dist_to_struc (Optional[float], optional): The maximum distance from a node to a structure. If not specified, calculation will not take it into account. Defaults to None.

        Raises:
            ValueError: Raised when any of the structure offsets, if specified, is smaller than zero.
            ValueError: Raised when any of the structure offsets, if specified, is greater than the branch length.
        """
        # Generate offsets
        self.branch_offsets = self._generate_offsets(
            mesh1d_edge_length, structure_chainage, max_dist_to_struc
        )
        # Calculate node positions
        self.node_xy = self.interpolate(self.branch_offsets)
        # Add mask (all False)
        self.mask = np.full(self.branch_offsets.shape, False)

    def _generate_offsets(
        self,
        mesh1d_edge_length: float,
        structure_offsets: Optional[List[float]] = None,
        max_dist_to_struc: Optional[float] = None,
    ) -> np.ndarray:
        """Generate the branch offsets.

        Args:
            mesh1d_edge_length (float): The edge length of the 1d mesh.
            structure_chainage (Optional[List[float]], optional): A list with the structure chainages. If not specified, calculation will not take it into account. Defaults to None.
            max_dist_to_struc (Optional[float], optional): The maximum distance from a node to a structure. If not specified, calculation will not take it into account. Defaults to None.

        Raises:
            ValueError: Raised when any of the structure offsets, if specified, is smaller than zero.
            ValueError: Raised when any of the structure offsets, if specified, is greater than the branch length.

        Returns:
            np.ndarray: The generated branch offsets.
        """
        # Generate initial offsets
        anchor_pts = [0.0, self.length]
        offsets = self._generate_1d_spacing(anchor_pts, mesh1d_edge_length)

        if structure_offsets is None:
            return offsets

        # Check the limits
        if (excess := min(structure_offsets)) < 0.0 or (
            excess := max(structure_offsets)
        ) > self.length:
            raise ValueError(
                f"Distance {excess} is outside the branch range (0.0 - {self.length})."
            )

        # Merge limits with start and end of branch
        limits = [-1e-3] + list(sorted(structure_offsets)) + [self.length + 1e-3]

        # if requested, check if the calculation point are close enough to the structures
        if max_dist_to_struc is not None:
            limits = self._generate_extended_limits(max_dist_to_struc, limits)

        offsets = self._add_nodes_to_segments(
            offsets, anchor_pts, limits, mesh1d_edge_length
        )

        return offsets

    def _generate_extended_limits(
        self, max_dist_to_struc: float, limits: List[float]
    ) -> List[float]:
        """Generate extended limits by taking into account the maximum distance to a structure.

        Args:
            max_dist_to_struc (float): The maximum distance from a node to a structure.
            limits (List[float]): The limits.

        Returns:
            List[float]: A list with the updated limits.
        """

        additional = []

        # Skip the first and the last, these are no structures
        for i in range(1, len(limits) - 1):
            # if the distance between two limits is large than twice the max distance to structure,
            # the mesh point will be too far away. Add a limit on the minimum of half the length and
            # two times the max distance
            dist_to_prev_limit = limits[i] - (
                max(additional[-1], limits[i - 1]) if any(additional) else limits[i - 1]
            )
            if dist_to_prev_limit > 2 * max_dist_to_struc:
                additional.append(
                    limits[i] - min(2 * max_dist_to_struc, dist_to_prev_limit / 2)
                )

            dist_to_next_limit = limits[i + 1] - limits[i]
            if dist_to_next_limit > 2 * max_dist_to_struc:
                additional.append(
                    limits[i] + min(2 * max_dist_to_struc, dist_to_next_limit / 2)
                )

        # Join the limits
        return sorted(limits + additional)

    def _add_nodes_to_segments(
        self,
        offsets: np.ndarray,
        anchor_pts: List[float],
        limits: List[float],
        mesh1d_edge_length: float,
    ) -> np.ndarray:
        """Add nodes to segments that are missing a mesh node.

        Args:
            offsets (np.ndarray): The branch offsets.
            anchor_pts (List[float]): The anchor points.
            limits (List[float]): The limits.
            mesh1d_edge_length (float): The edge length of the 1d mesh.

        Returns:
            np.ndarray: The array with branch offsets.
        """
        # Get upper and lower limits
        upper_limits = limits[1:]
        lower_limits = limits[:-1]

        def in_range():
            return [
                ((offsets > lower) & (offsets < upper)).any()
                for lower, upper in zip(lower_limits, upper_limits)
            ]

        # Determine the segments that are missing a mesh node
        # Anchor points are added on these segments, such that they will get a mesh node
        nodes_in_range = in_range()

        while not all(nodes_in_range):
            # Get the index of the first segment without grid point
            i = nodes_in_range.index(False)

            # Add it to the anchor pts
            anchor_pts.append((lower_limits[i] + upper_limits[i]) / 2.0)
            anchor_pts = sorted(anchor_pts)

            # Generate new offsets
            offsets = self._generate_1d_spacing(anchor_pts, mesh1d_edge_length)

            # Determine the segments that are missing a grid point
            nodes_in_range = in_range()

        if len(anchor_pts) > 2:
            logger.info(
                f"Added 1d mesh nodes on branch at: {anchor_pts}, due to the structures at {limits}."
            )

        return offsets

    @staticmethod
    def _generate_1d_spacing(
        anchor_pts: List[float], mesh1d_edge_length: float
    ) -> np.ndarray:
        """
        Generates 1d distances, called by function generate offsets
        """
        offsets = []
        # Loop through anchor point pairs
        for i in range(len(anchor_pts) - 1):
            # Determine section length between anchor point
            section_length = anchor_pts[i + 1] - anchor_pts[i]
            if section_length <= 0.0:
                raise ValueError("Section length must be larger than 0.0")
            # Determine number of nodes
            nnodes = max(2, int(round(section_length / mesh1d_edge_length) + 1)) - 1
            # Add nodes
            offsets.extend(
                np.linspace(
                    anchor_pts[i], anchor_pts[i + 1], nnodes, endpoint=False
                ).tolist()
            )
        # Add last node
        offsets.append(anchor_pts[-1])

        return np.asarray(offsets)

    def interpolate(self, distance: npt.ArrayLike) -> np.ndarray:
        """Interpolate coordinates along branch by length

        Args:
            distance (npt.ArrayLike): Length
        """
        intpcoords = np.stack(
            [
                np.interp(distance, self._distance, self._x_coordinates),
                np.interp(distance, self._distance, self._y_coordinates),
            ],
            axis=1,
        )

        return intpcoords

generate_nodes(mesh1d_edge_length, structure_chainage=None, max_dist_to_struc=None)

Generate the branch offsets and the nodes.

Parameters:

Name	Type	Description	Default
mesh1d_edge_length	float	The edge length of the 1d mesh.	required
structure_chainage	Optional[List[float]]	A list with the structure chainages. If not specified, calculation will not take it into account. Defaults to None.	None
max_dist_to_struc	Optional[float]	The maximum distance from a node to a structure. If not specified, calculation will not take it into account. Defaults to None.	None

Raises:

Type	Description
ValueError	Raised when any of the structure offsets, if specified, is smaller than zero.
ValueError	Raised when any of the structure offsets, if specified, is greater than the branch length.

Source code in hydrolib/core/dflowfm/net/models.py

def generate_nodes(
    self,
    mesh1d_edge_length: float,
    structure_chainage: Optional[List[float]] = None,
    max_dist_to_struc: Optional[float] = None,
):
    """Generate the branch offsets and the nodes.

    Args:
        mesh1d_edge_length (float): The edge length of the 1d mesh.
        structure_chainage (Optional[List[float]], optional): A list with the structure chainages. If not specified, calculation will not take it into account. Defaults to None.
        max_dist_to_struc (Optional[float], optional): The maximum distance from a node to a structure. If not specified, calculation will not take it into account. Defaults to None.

    Raises:
        ValueError: Raised when any of the structure offsets, if specified, is smaller than zero.
        ValueError: Raised when any of the structure offsets, if specified, is greater than the branch length.
    """
    # Generate offsets
    self.branch_offsets = self._generate_offsets(
        mesh1d_edge_length, structure_chainage, max_dist_to_struc
    )
    # Calculate node positions
    self.node_xy = self.interpolate(self.branch_offsets)
    # Add mask (all False)
    self.mask = np.full(self.branch_offsets.shape, False)

interpolate(distance)

Interpolate coordinates along branch by length

Parameters:

Name	Type	Description	Default
distance	ArrayLike	Length	required

Source code in hydrolib/core/dflowfm/net/models.py

def interpolate(self, distance: npt.ArrayLike) -> np.ndarray:
    """Interpolate coordinates along branch by length

    Args:
        distance (npt.ArrayLike): Length
    """
    intpcoords = np.stack(
        [
            np.interp(distance, self._distance, self._x_coordinates),
            np.interp(distance, self._distance, self._y_coordinates),
        ],
        axis=1,
    )

    return intpcoords

Link1d2d

Bases: BaseModel

Link1d2d defines the 1D2D Links of a model network.

Attributes:

Name	Type	Description
meshkernel	Optional[MeshKernel]	The MeshKernel used to interact with this Link1d2d
link1d2d_id	ndarray	The id of this Link1d2d
link1d2d_long_name	ndarray	The long name of this Link1d2d
link1d2d_contact_type	ndarray	The contact type of this Link1d2d
link1d2d	ndarray	The underlying data object of this Link1d2d

Source code in hydrolib/core/dflowfm/net/models.py

class Link1d2d(BaseModel):
    """Link1d2d defines the 1D2D Links of a model network.

    Attributes:
        meshkernel (Optional[mk.MeshKernel]):
            The MeshKernel used to interact with this Link1d2d
        link1d2d_id (np.ndarray):
            The id of this Link1d2d
        link1d2d_long_name (np.ndarray):
            The long name of this Link1d2d
        link1d2d_contact_type (np.ndarray):
            The contact type of this Link1d2d
        link1d2d (np.ndarray):
            The underlying data object of this Link1d2d
    """

    meshkernel: mk.MeshKernel = Field(default_factory=mk.MeshKernel)

    @property
    def link1d2d(self) -> np.ndarray[np.int32]:
        contacts = self.meshkernel.contacts_get()
        link1d2d_arr = np.stack(
            [contacts.mesh1d_indices, contacts.mesh2d_indices], axis=1
        )
        return link1d2d_arr

    @property
    def link1d2d_contact_type(self) -> np.ndarray[np.int32]:
        contacts = self.meshkernel.contacts_get()
        link1d2d_contact_type_arr = np.full(contacts.mesh1d_indices.size, 3)
        return link1d2d_contact_type_arr

    @property
    def link1d2d_id(self) -> np.ndarray[object]:
        link1d2d_id_arr = np.array([f"{n1d:d}_{f2d:d}" for n1d, f2d in self.link1d2d])
        return link1d2d_id_arr

    @property
    def link1d2d_long_name(self) -> np.ndarray[object]:
        link1d2d_id_arr = np.array([f"{n1d:d}_{f2d:d}" for n1d, f2d in self.link1d2d])
        return link1d2d_id_arr

    def is_empty(self) -> bool:
        """Whether this Link1d2d is currently empty.

        Returns:
            bool: True if the Link1d2d is currently empty; False otherwise.
        """
        return self.link1d2d.size == 0

    def read_file(self, file_path: Path) -> None:
        """Read the Link1d2d data from the specified netCDF file at file_path into this

        Args:
            file_path (Path): Path to the netCDF file.
        """

        reader = UgridReader(file_path)
        reader.read_link1d2d(self)

    def clear(self) -> None:
        """Remove all saved links from the links administration"""
        self.link1d2d_id = np.empty(0, object)
        self.link1d2d_long_name = np.empty(0, object)
        self.link1d2d_contact_type = np.empty(0, np.int32)
        self.link1d2d = np.empty((0, 2), np.int32)
        # The meshkernel object needs to be resetted
        self.meshkernel._deallocate_state()
        self.meshkernel._allocate_state(self.meshkernel.get_projection())
        self.meshkernel.contacts_get()

    def _link_from_1d_to_2d(
        self, node_mask: np.ndarray, polygon: mk.GeometryList = None
    ):
        """Connect 1d nodes to 2d face circumcenters. A list of branchid's can be given
        to indicate where the 1d-side of the connections should be made. A polygon can
        be given to indicate where the 2d-side of the connections should be made.

        Note that the links are added to the already existing links. To remove these, use the method "clear".

        Args:
            node_mask (np.ndarray): Array indicating what 1d nodes should be connected. Defaults to None.
            polygon (mk.GeometryList): Coordinates of the area within which the 2d side of the links are connected.
        """

        # Computes Mesh1d-Mesh2d contacts, where each single Mesh1d node is connected to one Mesh2d face circumcenter.
        # The boundary nodes of Mesh1d (those sharing only one Mesh1d edge) are not connected to any Mesh2d face.
        self.meshkernel.contacts_compute_single(
            node_mask=node_mask, polygons=polygon, projection_factor=1.0
        )

        # Note that the function "contacts_compute_multiple" also computes the connections, but does not take into account
        # a bounding polygon or the end points of the 1d mesh.

    def _link_from_2d_to_1d_embedded(
        self, node_mask: np.ndarray, polygons: mk.GeometryList
    ):
        """"""
        self.meshkernel.contacts_compute_with_points(
            node_mask=node_mask, polygons=polygons
        )

    def _link_from_2d_to_1d_lateral(
        self,
        node_mask: np.ndarray,
        # boundary_face_xy: np.ndarray,
        polygon: mk.GeometryList = None,
        search_radius: float = None,
    ):
        # TODO: Missing value double for search radius?

        # Computes Mesh1d-Mesh2d contacts, where Mesh1d nodes are connected to the closest Mesh2d faces at the boundary
        self.meshkernel.contacts_compute_boundary(
            node_mask=node_mask, polygons=polygon, search_radius=search_radius
        )

clear()

Remove all saved links from the links administration

Source code in hydrolib/core/dflowfm/net/models.py

def clear(self) -> None:
    """Remove all saved links from the links administration"""
    self.link1d2d_id = np.empty(0, object)
    self.link1d2d_long_name = np.empty(0, object)
    self.link1d2d_contact_type = np.empty(0, np.int32)
    self.link1d2d = np.empty((0, 2), np.int32)
    # The meshkernel object needs to be resetted
    self.meshkernel._deallocate_state()
    self.meshkernel._allocate_state(self.meshkernel.get_projection())
    self.meshkernel.contacts_get()

is_empty()

Whether this Link1d2d is currently empty.

Returns:

Name	Type	Description
bool	bool	True if the Link1d2d is currently empty; False otherwise.

Source code in hydrolib/core/dflowfm/net/models.py

def is_empty(self) -> bool:
    """Whether this Link1d2d is currently empty.

    Returns:
        bool: True if the Link1d2d is currently empty; False otherwise.
    """
    return self.link1d2d.size == 0

read_file(file_path)

Read the Link1d2d data from the specified netCDF file at file_path into this

Parameters:

Name	Type	Description	Default
file_path	Path	Path to the netCDF file.	required

Source code in hydrolib/core/dflowfm/net/models.py

def read_file(self, file_path: Path) -> None:
    """Read the Link1d2d data from the specified netCDF file at file_path into this

    Args:
        file_path (Path): Path to the netCDF file.
    """

    reader = UgridReader(file_path)
    reader.read_link1d2d(self)

Mesh1d

Bases: BaseModel

Source code in hydrolib/core/dflowfm/net/models.py

class Mesh1d(BaseModel):
    """"""

    meshkernel: mk.MeshKernel = Field(default_factory=mk.MeshKernel)

    branches: Dict[str, Branch] = {}

    network1d_node_id: np.ndarray = Field(default_factory=lambda: np.empty(0, object))
    network1d_node_long_name: np.ndarray = Field(
        default_factory=lambda: np.empty(0, object)
    )
    network1d_node_x: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    network1d_node_y: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    network1d_branch_id: np.ndarray = Field(default_factory=lambda: np.empty(0, object))
    network1d_branch_long_name: np.ndarray = Field(
        default_factory=lambda: np.empty(0, object)
    )
    network1d_branch_length: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.double)
    )
    network1d_branch_order: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.int32)
    )
    network1d_edge_nodes: np.ndarray = Field(
        default_factory=lambda: np.empty((0, 2), np.int32)
    )
    # TODO: sync with node_x/node_y/edge_nodes with meshkernel: https://github.com/Deltares/HYDROLIB-core/issues/576
    network1d_geom_x: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    network1d_geom_y: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    network1d_part_node_count: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.int32)
    )

    mesh1d_node_x: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    mesh1d_node_y: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    mesh1d_node_id: np.ndarray = Field(default_factory=lambda: np.empty(0, object))
    mesh1d_node_long_name: np.ndarray = Field(
        default_factory=lambda: np.empty(0, object)
    )
    mesh1d_node_branch_id: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.int32)
    )
    mesh1d_node_branch_offset: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.double)
    )

    mesh1d_edge_nodes: np.ndarray = Field(
        default_factory=lambda: np.empty((0, 2), np.int32)
    )
    mesh1d_edge_x: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    mesh1d_edge_y: np.ndarray = Field(default_factory=lambda: np.empty(0, np.double))
    mesh1d_edge_branch_id: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.int32)
    )
    mesh1d_edge_branch_offset: np.ndarray = Field(
        default_factory=lambda: np.empty(0, np.double)
    )

    def is_empty(self) -> bool:
        return self.mesh1d_node_x.size == 0

    def _get_mesh1d(self) -> mk.Mesh1d:
        """Return mesh1d from meshkernel. Note that the meshkernel.Mesh1d instance
        does not contain all mesh attributes that are contained in this class"""
        return self.meshkernel.mesh1d_get()

    def _set_mesh1d(self) -> None:
        self.meshkernel.mesh1d_set(
            mk.Mesh1d(
                node_x=self.mesh1d_node_x.astype(np.float64),
                node_y=self.mesh1d_node_y.astype(np.float64),
                edge_nodes=self.mesh1d_edge_nodes.ravel().astype(np.int32),
            )
        )

    def _process_network1d(self) -> None:
        """
        Determine x, y locations of mesh1d nodes based on the network1d
        """
        # Create a list of coordinates to create the branches from
        ngeom = list(zip(self.network1d_geom_x, self.network1d_geom_y))

        self.branches.clear()

        for i, (name, nnodes) in enumerate(
            zip(self.network1d_branch_id, self.network1d_part_node_count)
        ):

            # Create network branch
            # Get geometry of branch from network geometry
            geometry = np.array([ngeom.pop(0) for _ in range(nnodes)])
            # Get branch offsets
            idx = self.mesh1d_node_branch_id == i
            branch_offsets = self.mesh1d_node_branch_offset[idx]
            mask = np.full(branch_offsets.shape, False)

            # Determine if a start or end coordinate needs to be added for constructing a complete branch
            # As nodes are re-used, the last and first branch_offsets are often missing. However, they are still used
            # for determining the length along the discretized branch.
            if branch_offsets.size == 0 or not np.isclose(branch_offsets[0], 0.0):
                branch_offsets = np.concatenate([[0], branch_offsets])
                mask = np.concatenate([[True], mask])
            length = np.hypot(*np.diff(geometry, axis=0).T).sum()
            if not np.isclose(branch_offsets[-1], length):
                branch_offsets = np.concatenate([branch_offsets, [length]])
                mask = np.concatenate([mask, [True]])

            # Create instance of branch object and add to dictionary
            geo_branch = Branch(geometry, branch_offsets=branch_offsets, mask=mask)
            self.branches[name.strip()] = geo_branch

        # Convert list with all coordinates (except the appended ones for the schematized branches) to arrays
        node_x, node_y = np.vstack(
            [branch.node_xy[~branch.mask] for branch in self.branches.values()]
        ).T

        # Add to variables
        self.mesh1d_node_x = node_x
        self.mesh1d_node_y = node_y

        # Calculate edge coordinates
        edge_x, edge_y = np.vstack(
            [
                branch.interpolate(
                    self.mesh1d_edge_branch_offset[self.mesh1d_edge_branch_id == i]
                )
                for i, branch in enumerate(self.branches.values())
            ]
        ).T

        # Add to variables
        self.mesh1d_edge_x = edge_x
        self.mesh1d_edge_y = edge_y

    def _network1d_node_position(self, x: float, y: float) -> Union[np.int32, None]:
        """Determine the position (index) of a x, y coordinate in the network nodes

        Args:
            x (float): x-coordinate
            y (float): y-coordinate

        Returns:
            Union[np.int32, None]: The index of the coordinate. None if not found
        """
        return self._node_position(self.network1d_node_x, self.network1d_node_y, x, y)

    def _mesh1d_node_position(self, x: float, y: float) -> Union[np.int32, None]:
        """Determine the position (index) of a x, y coordinate in the mesh nodes

        Args:
            x (float): x-coordinate
            y (float): y-coordinate

        Returns:
            Union[np.int32, None]: The index of the coordinate. None if not found
        """
        return self._node_position(self.mesh1d_node_x, self.mesh1d_node_y, x, y)

    def _node_position(
        self, arrx: np.ndarray, arry: np.ndarray, x: float, y: float
    ) -> Union[np.int32, None]:
        """Determine the position (index) of a x, y coordinate in a given x and y array

        Args:
            arrx (np.ndarray): x-coordinates in which the position is sought
            arry (np.ndarray): y-coordiantes in which the position is sought
            x (float): x-coordinate to be sought
            y (float): y-coordinate to be sought

        Raises:
            ValueError: If multiple positions are found for the coordinate

        Returns:
            Union[np.int32, None]: The index of the coordinate. None if not found
        """
        pos = np.where(np.isclose(arrx, x, rtol=0.0) & np.isclose(arry, y, rtol=0.0))[0]
        if pos.size == 0:
            return None
        elif pos.size == 1:
            return np.int32(pos[0])
        else:
            # Find the nearest
            distance = np.hypot(arrx[pos] - x, arry[pos] - y)
            if np.unique(distance).size == 1:
                raise ValueError("Multiple nodes were found at the same position.")
            else:
                return np.int32(pos[np.argmin(distance)])

    def _add_branch(
        self,
        branch: Branch,
        name: str = None,
        branch_order: int = -1,
        long_name: str = None,
        force_midpoint: bool = True,
    ):
        """Add the branch to mesh1d

        Args:
            branch (Branch): branch to add to the mesh1d
            name (str): id of the branch
            branch_order (int): interpolation order of the branch
            long_name (str): long name of the branch
            force_midpoint(bool): argument to control if a midpoint will be forced on the branch, use False for pipes

        Returns:
            Str: name of the branch.
        """

        # Check if branch had coordinate discretization
        if branch.branch_offsets.size == 0:
            raise ValueError(
                'Branch has no mesh discretization. Use the function "generate_nodes" solve generate a 1d mesh on the branch.'
            )

        if name in self.network1d_branch_id:
            raise KeyError(f'The branch name "{name}" is already used.')
        if long_name in self.network1d_branch_long_name:
            raise KeyError(f'The branch long name "{long_name}" is already used.')

        branch_nr = len(self.network1d_branch_id)
        if name is None:
            name = f"br{branch_nr:05d}"
        if long_name is None:
            long_name = name

        self.branches[name] = branch

        # Add branch administration
        self.network1d_branch_order = np.append(
            self.network1d_branch_order, branch_order
        )
        self.network1d_branch_length = np.append(
            self.network1d_branch_length, branch.length
        )
        self.network1d_branch_id = np.append(self.network1d_branch_id, name)
        self.network1d_branch_long_name = np.append(
            self.network1d_branch_long_name, long_name
        )

        # Add branch geometry coordinates
        self.network1d_part_node_count = np.append(
            self.network1d_part_node_count, len(branch.geometry)
        )
        self.network1d_geom_x = np.append(self.network1d_geom_x, branch._x_coordinates)
        self.network1d_geom_y = np.append(self.network1d_geom_y, branch._y_coordinates)

        # Network edge node administration
        # -------------------------------

        first_point = branch.geometry[0]
        last_point = branch.geometry[-1]

        # Get offsets from dictionary
        offsets = branch.branch_offsets[:]
        # The number of links on the branch
        nlinks = len(offsets) - 1

        # Check if the first and last point of the branch are already in the set
        first_present = self._network1d_node_position(*first_point) is not None
        if first_present:
            # If present, remove from branch offsets
            offsets = offsets[1:]
            branch.mask[0] = True
        else:
            # If not present, add to network nodes
            self.network1d_node_x = np.append(self.network1d_node_x, first_point[0])
            self.network1d_node_y = np.append(self.network1d_node_y, first_point[1])

            self.network1d_node_id = np.append(
                self.network1d_node_id, "{:.6f}_{:.6f}".format(*first_point)
            )
            self.network1d_node_long_name = np.append(
                self.network1d_node_long_name, "x={:.6f}_y={:.6f}".format(*first_point)
            )

        last_present = self._network1d_node_position(*last_point) is not None
        if last_present:
            # If present, remove from branch offsets
            offsets = offsets[:-1]
            branch.mask[-1] = True
        else:
            # If not present, add to network nodes
            self.network1d_node_x = np.append(self.network1d_node_x, last_point[0])
            self.network1d_node_y = np.append(self.network1d_node_y, last_point[1])

            self.network1d_node_id = np.append(
                self.network1d_node_id, "{:.6f}_{:.6f}".format(*last_point)
            )
            self.network1d_node_long_name = np.append(
                self.network1d_node_long_name, "x={:.6f}_y={:.6f}".format(*last_point)
            )

        # If no points remain, add an extra halfway: each branch should have at least 1 node
        # Adjust the branch object as well, by adding the extra point
        if len(offsets) == 0 and force_midpoint:
            # Add extra offset
            extra_offset = branch.length / 2.0
            offsets = np.array([extra_offset])
            nlinks += 1
            # Adjust branch object
            branch.branch_offsets = np.insert(branch.branch_offsets, 1, extra_offset)
            branch.node_xy = np.insert(
                branch.node_xy, 1, branch.interpolate(offsets), axis=0
            )
            branch.mask = np.insert(branch.mask, 1, False)

        # Get the index of the first and last node, add as edge_nodes
        i_from = self._network1d_node_position(first_point[0], first_point[1])
        i_to = self._network1d_node_position(last_point[0], last_point[1])
        if i_from == i_to:
            raise ValueError(
                "Start and end node are the same. Ring geometries are not accepted."
            )

        self.network1d_edge_nodes = np.append(
            self.network1d_edge_nodes,
            np.array([[i_from, i_to]], dtype=np.int32),
            axis=0,
        )

        # Mesh1d edge node administration

        # -------------------------------
        # First determine the start index. This is equal to the number of already present points
        start_index = len(self.mesh1d_node_branch_id)
        # For each link, create a new edge node connection
        # If the first node is already present, subtract 1, since the first number will be substitud with the present node
        if first_present:
            start_index -= 1
        new_edge_nodes = (
            np.stack([np.arange(nlinks), np.arange(nlinks) + 1], axis=1) + start_index
        ).astype(np.int32)

        # If the first node is present, change the first point of the first edge to the existing point
        if first_present:
            new_edge_nodes[0, 0] = self._mesh1d_node_position(*first_point)
        # If the last node is present, change the last point of the last edge too
        if last_present:
            new_edge_nodes[-1, 1] = self._mesh1d_node_position(*last_point)

        # Add to variables
        self.mesh1d_node_x = np.append(
            self.mesh1d_node_x, branch.node_xy[~branch.mask, 0]
        )
        self.mesh1d_node_y = np.append(
            self.mesh1d_node_y, branch.node_xy[~branch.mask, 1]
        )

        # Add to edge_nodes
        self.mesh1d_edge_nodes = np.append(
            self.mesh1d_edge_nodes, new_edge_nodes, axis=0
        )
        edge_coords = np.stack([self.mesh1d_node_x, self.mesh1d_node_y], axis=1)[
            new_edge_nodes
        ].mean(1)
        edge_offsets = (branch.branch_offsets[:-1] + branch.branch_offsets[1:]) / 2

        self.mesh1d_edge_branch_id = np.append(
            self.mesh1d_edge_branch_id, np.full(len(edge_coords), branch_nr)
        )
        self.mesh1d_edge_branch_offset = np.append(
            self.mesh1d_edge_branch_offset, edge_offsets
        )

        self.mesh1d_edge_x = np.append(self.mesh1d_edge_x, edge_coords[:, 0])
        self.mesh1d_edge_y = np.append(self.mesh1d_edge_y, edge_coords[:, 1])

        # Update names of nodes
        mesh_point_names = np.array(
            [f"{name}_{offset:.2f}" for offset in offsets], dtype=object
        )
        self.mesh1d_node_id = np.append(self.mesh1d_node_id, mesh_point_names)
        self.mesh1d_node_long_name = np.append(
            self.mesh1d_node_long_name, mesh_point_names
        )

        # Add mesh1d nodes
        self.mesh1d_node_branch_id = np.append(
            self.mesh1d_node_branch_id, np.full(len(offsets), branch_nr)
        )
        self.mesh1d_node_branch_offset = np.append(
            self.mesh1d_node_branch_offset, offsets
        )
        return name

    def get_node_mask(self, branchids: List[str] = None):
        """Get node mask, give a mask with True for each node that is in the given branchid list"""

        mask = np.full(self.mesh1d_node_id.shape, False, dtype=bool)
        if branchids is None:
            mask[:] = True
            return mask

        # Get number (index) of given branches
        idx = np.where(np.isin(self.network1d_branch_id, branchids))[0]
        if idx.size == 0:
            raise KeyError("No branches corresponding to the given keys were found.")

        mask[np.isin(self.mesh1d_node_branch_id, idx)] = True

        return mask

get_node_mask(branchids=None)

Get node mask, give a mask with True for each node that is in the given branchid list

Source code in hydrolib/core/dflowfm/net/models.py

def get_node_mask(self, branchids: List[str] = None):
    """Get node mask, give a mask with True for each node that is in the given branchid list"""

    mask = np.full(self.mesh1d_node_id.shape, False, dtype=bool)
    if branchids is None:
        mask[:] = True
        return mask

    # Get number (index) of given branches
    idx = np.where(np.isin(self.network1d_branch_id, branchids))[0]
    if idx.size == 0:
        raise KeyError("No branches corresponding to the given keys were found.")

    mask[np.isin(self.mesh1d_node_branch_id, idx)] = True

    return mask

Mesh2d

Bases: BaseModel

Mesh2d defines a single two dimensional grid.

Attributes:

Name	Type	Description
meshkernel	MeshKernel	The meshkernel used to manimpulate this Mesh2d.
mesh2d_node_z	ndarray	The node positions on the z-axis. Defaults to np.empty(0, dtype=np.double).
mesh2d_face_z	ndarray	The face positions on the z-axis. Defaults to np.empty(0, dtype=np.double).

Source code in hydrolib/core/dflowfm/net/models.py

class Mesh2d(BaseModel):
    """Mesh2d defines a single two dimensional grid.

    Attributes:
        meshkernel (mk.MeshKernel):
            The meshkernel used to manimpulate this Mesh2d.
        mesh2d_node_z (np.ndarray):
            The node positions on the z-axis. Defaults to np.empty(0, dtype=np.double).
        mesh2d_face_z (np.ndarray):
            The face positions on the z-axis. Defaults to np.empty(0, dtype=np.double).
    """

    meshkernel: mk.MeshKernel = Field(default_factory=mk.MeshKernel)

    # placeholders for bathymetry
    mesh2d_node_z: np.ndarray = Field(
        default_factory=lambda: np.empty(0, dtype=np.double)
    )
    mesh2d_face_z: np.ndarray = Field(
        default_factory=lambda: np.empty(0, dtype=np.double)
    )

    @property
    def mesh2d_node_x(self) -> np.ndarray[float]:
        """The x-coordinates of the nodes in the mesh.

        Returns:
            ndarray[float]: A 1D double array describing the x-coordinates of the nodes.
        """
        return self.meshkernel.mesh2d_get().node_x

    @property
    def mesh2d_node_y(self) -> np.ndarray[float]:
        """The y-coordinates of the nodes in the mesh.

        Returns:
            ndarray[float]: A 1D double array describing the y-coordinates of the nodes.
        """
        return self.meshkernel.mesh2d_get().node_y

    @property
    def mesh2d_edge_x(self) -> np.ndarray[float]:
        """The x-coordinates of the mesh edges' middle points.

        Returns:
            ndarray[float]: A 1D double array describing x-coordinates of the mesh edges' middle points.
        """
        return self.meshkernel.mesh2d_get().edge_x

    @property
    def mesh2d_edge_y(self) -> np.ndarray[float]:
        """The y-coordinates of the mesh edges' middle points.

        Returns:
            ndarray[float]: A 1D double array describing y-coordinates of the mesh edges' middle points.
        """
        return self.meshkernel.mesh2d_get().edge_y

    @property
    def mesh2d_edge_nodes(self) -> np.ndarray[int, int]:
        """The node indices of the mesh edges.

        Returns:
            np.ndarray[int, int]: A 2D integer array (nEdges, 2) containg the two node indices for each edge.
        """
        mesh2d_output = self.meshkernel.mesh2d_get()
        edge_nodes = mesh2d_output.edge_nodes.reshape((-1, 2))
        return edge_nodes

    @property
    def mesh2d_face_x(self) -> np.ndarray[float]:
        """The x-coordinates of the mesh faces' mass centers.

        Returns:
            ndarray[float]: A 1D double array describing x-coordinates of the mesh faces' mass centers.
        """
        return self.meshkernel.mesh2d_get().face_x

    @property
    def mesh2d_face_y(self) -> np.ndarray[float]:
        """The y-coordinates of the mesh faces' mass centers.

        Returns:
            ndarray[float]: A 1D double array describing y-coordinates of the mesh faces' mass centers.
        """
        return self.meshkernel.mesh2d_get().face_y

    @property
    def mesh2d_face_nodes(self) -> np.ndarray[int, int]:
        """The node indices of the mesh faces

        Returns:
            np.ndarray[int, int]: A 2D integer array describing the nodes composing each mesh 2d face. A 2D integer array (nFaces, maxNodesPerFace) containg the node indices for each face.
        """
        mesh2d_output = self.meshkernel.mesh2d_get()
        npf = mesh2d_output.nodes_per_face
        if self.is_empty():
            return np.empty((0, 0), dtype=np.int32)
        face_node_connectivity = np.full(
            (len(mesh2d_output.face_x), max(npf)), np.iinfo(np.int32).min
        )
        idx = (
            np.ones_like(face_node_connectivity) * np.arange(max(npf))[None, :]
        ) < npf[:, None]
        face_node_connectivity[idx] = mesh2d_output.face_nodes
        return face_node_connectivity

    def is_empty(self) -> bool:
        """Determine whether this Mesh2d is empty.

        Returns:
            (bool): Whether this Mesh2d is empty.
        """
        return self.mesh2d_node_x.size == 0

    def read_file(self, file_path: Path) -> None:
        """Read the Mesh2d from the file at file_path.

        Args:
            file_path (Path): Path to the file to be read.
        """
        reader = UgridReader(file_path)
        reader.read_mesh2d(self)

    def _set_mesh2d(self, node_x, node_y, edge_nodes) -> None:
        mesh2d = mk.Mesh2d(
            node_x=node_x.astype(np.float64),
            node_y=node_y.astype(np.float64),
            edge_nodes=edge_nodes.ravel().astype(np.int32),
        )

        self.meshkernel.mesh2d_set(mesh2d)

    def get_mesh2d(self) -> mk.Mesh2d:
        """Get the mesh2d as represented in the MeshKernel

        Returns:
            (mk.Mesh2d): The mesh2d as represented in the MeshKernel
        """
        return self.meshkernel.mesh2d_get()

    def create_rectilinear(self, extent: tuple, dx: float, dy: float) -> None:
        """Create a rectilinear mesh within a polygon. A rectangular grid is generated within the polygon bounds

        Args:
            extent (tuple): Bounding box of mesh (left, bottom, right, top)
            dx (float): Horizontal distance
            dy (float): Vertical distance

        Raises:
            NotImplementedError: MultiPolygons
        """

        xmin, ymin, xmax, ymax = extent

        rows = int((ymax - ymin) / dy)
        columns = int((xmax - xmin) / dx)

        params = mk.MakeGridParameters(
            num_columns=columns,
            num_rows=rows,
            origin_x=xmin,
            origin_y=ymin,
            block_size_x=dx,
            block_size_y=dy,
        )

        mesh2d_input = self.meshkernel  # mk.MeshKernel()
        mesh2d_input.curvilinear_compute_rectangular_grid(params)
        mesh2d_input.curvilinear_convert_to_mesh2d()  # convert to ugrid/mesh2d

    def create_triangular(self, geometry_list: mk.GeometryList) -> None:
        """Create triangular grid within GeometryList object

        Args:
            geometry_list (mk.GeometryList): GeometryList represeting a polygon within which the mesh is generated.
        """
        # Call meshkernel
        self.meshkernel.mesh2d_make_triangular_mesh_from_polygon(geometry_list)

    def clip(
        self,
        geometrylist: mk.GeometryList,
        deletemeshoption: mk.DeleteMeshOption = mk.DeleteMeshOption.INSIDE_NOT_INTERSECTED,
        inside=False,
    ) -> None:
        """Clip the 2D mesh by a polygon. Both outside the exterior and inside the interiors is clipped. It is also possible to clip inside a polygon with holes.


        Args:
            geometrylist (GeometryList): Polygon stored as GeometryList
            deletemeshoption (int, optional): [description]. Defaults to 1.
        """

        # For clipping outside
        if not inside:
            # Check if a multipolygon was provided when clipping outside
            if geometrylist.geometry_separator in geometrylist.x_coordinates:
                raise NotImplementedError(
                    "Deleting outside more than a single exterior (MultiPolygon) is not implemented."
                )

            # Get exterior and interiors
            parts = split_by(geometrylist, geometrylist.inner_outer_separator)

            exteriors = [parts[0]]
            interiors = parts[1:]

            # Check if parts are closed
            for part in exteriors + interiors:
                if (part.x_coordinates[0], part.y_coordinates[0]) != (
                    part.x_coordinates[-1],
                    part.y_coordinates[-1],
                ):
                    raise ValueError(
                        "First and last coordinate of each GeometryList part should match."
                    )

        # Inside
        else:
            # Get exterior and interiors
            parts = split_by(geometrylist, geometrylist.geometry_separator)
            exteriors = parts[:]
            interiors = []

        # Delete everything outside the (Multi)Polygon
        for exterior in exteriors:
            self.meshkernel.mesh2d_delete(
                geometry_list=exterior,
                delete_option=deletemeshoption,
                invert_deletion=not inside,
            )

        # Delete all holes.
        for interior in interiors:
            self.meshkernel.mesh2d_delete(
                geometry_list=interior,
                delete_option=deletemeshoption,
                invert_deletion=inside,
            )

    def refine(
        self,
        polygon: mk.GeometryList,
        level: int,
        parameters: mk.MeshRefinementParameters = None,
    ):
        """Refine the mesh within a polygon, by a number of steps (level)

        Args:
            polygon (GeometryList): Polygon in which to refine
            level (int): Number of refinement steps
            parameters (MeshRefinementParameters): object containing the Meshkernel refinement parameters. Default values are taken from meskernel documentation.
        """
        if parameters is None:
            parameters = mk.MeshRefinementParameters(
                refine_intersected=True,
                use_mass_center_when_refining=True,
                min_edge_size=1.0,
                refinement_type=2,
                connect_hanging_nodes=True,
                account_for_samples_outside_face=True,
                max_refinement_iterations=1,
                smoothing_iterations=5,
                max_courant_time=120.0,
                directional_refinement=False,
            )
        parameters.max_refinement_iterations = level
        self.meshkernel.mesh2d_refine_based_on_polygon(polygon, parameters)

mesh2d_edge_nodes property

The node indices of the mesh edges.

Returns:

Type	Description
ndarray[int, int]	np.ndarray[int, int]: A 2D integer array (nEdges, 2) containg the two node indices for each edge.

mesh2d_edge_x property

The x-coordinates of the mesh edges' middle points.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing x-coordinates of the mesh edges' middle points.

mesh2d_edge_y property

The y-coordinates of the mesh edges' middle points.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing y-coordinates of the mesh edges' middle points.

mesh2d_face_nodes property

The node indices of the mesh faces

Returns:

Type	Description
ndarray[int, int]	np.ndarray[int, int]: A 2D integer array describing the nodes composing each mesh 2d face. A 2D integer array (nFaces, maxNodesPerFace) containg the node indices for each face.

mesh2d_face_x property

The x-coordinates of the mesh faces' mass centers.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing x-coordinates of the mesh faces' mass centers.

mesh2d_face_y property

The y-coordinates of the mesh faces' mass centers.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing y-coordinates of the mesh faces' mass centers.

mesh2d_node_x property

The x-coordinates of the nodes in the mesh.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing the x-coordinates of the nodes.

mesh2d_node_y property

The y-coordinates of the nodes in the mesh.

Returns:

Type	Description
ndarray[float]	ndarray[float]: A 1D double array describing the y-coordinates of the nodes.

clip(geometrylist, deletemeshoption=mk.DeleteMeshOption.INSIDE_NOT_INTERSECTED, inside=False)

Clip the 2D mesh by a polygon. Both outside the exterior and inside the interiors is clipped. It is also possible to clip inside a polygon with holes.

Parameters:

Name	Type	Description	Default
geometrylist	GeometryList	Polygon stored as GeometryList	required
deletemeshoption	int	[description]. Defaults to 1.	INSIDE_NOT_INTERSECTED

Source code in hydrolib/core/dflowfm/net/models.py

def clip(
    self,
    geometrylist: mk.GeometryList,
    deletemeshoption: mk.DeleteMeshOption = mk.DeleteMeshOption.INSIDE_NOT_INTERSECTED,
    inside=False,
) -> None:
    """Clip the 2D mesh by a polygon. Both outside the exterior and inside the interiors is clipped. It is also possible to clip inside a polygon with holes.


    Args:
        geometrylist (GeometryList): Polygon stored as GeometryList
        deletemeshoption (int, optional): [description]. Defaults to 1.
    """

    # For clipping outside
    if not inside:
        # Check if a multipolygon was provided when clipping outside
        if geometrylist.geometry_separator in geometrylist.x_coordinates:
            raise NotImplementedError(
                "Deleting outside more than a single exterior (MultiPolygon) is not implemented."
            )

        # Get exterior and interiors
        parts = split_by(geometrylist, geometrylist.inner_outer_separator)

        exteriors = [parts[0]]
        interiors = parts[1:]

        # Check if parts are closed
        for part in exteriors + interiors:
            if (part.x_coordinates[0], part.y_coordinates[0]) != (
                part.x_coordinates[-1],
                part.y_coordinates[-1],
            ):
                raise ValueError(
                    "First and last coordinate of each GeometryList part should match."
                )

    # Inside
    else:
        # Get exterior and interiors
        parts = split_by(geometrylist, geometrylist.geometry_separator)
        exteriors = parts[:]
        interiors = []

    # Delete everything outside the (Multi)Polygon
    for exterior in exteriors:
        self.meshkernel.mesh2d_delete(
            geometry_list=exterior,
            delete_option=deletemeshoption,
            invert_deletion=not inside,
        )

    # Delete all holes.
    for interior in interiors:
        self.meshkernel.mesh2d_delete(
            geometry_list=interior,
            delete_option=deletemeshoption,
            invert_deletion=inside,
        )

create_rectilinear(extent, dx, dy)

Create a rectilinear mesh within a polygon. A rectangular grid is generated within the polygon bounds

Parameters:

Name	Type	Description	Default
extent	tuple	Bounding box of mesh (left, bottom, right, top)	required
dx	float	Horizontal distance	required
dy	float	Vertical distance	required

Raises:

Type	Description
NotImplementedError	MultiPolygons

Source code in hydrolib/core/dflowfm/net/models.py

def create_rectilinear(self, extent: tuple, dx: float, dy: float) -> None:
    """Create a rectilinear mesh within a polygon. A rectangular grid is generated within the polygon bounds

    Args:
        extent (tuple): Bounding box of mesh (left, bottom, right, top)
        dx (float): Horizontal distance
        dy (float): Vertical distance

    Raises:
        NotImplementedError: MultiPolygons
    """

    xmin, ymin, xmax, ymax = extent

    rows = int((ymax - ymin) / dy)
    columns = int((xmax - xmin) / dx)

    params = mk.MakeGridParameters(
        num_columns=columns,
        num_rows=rows,
        origin_x=xmin,
        origin_y=ymin,
        block_size_x=dx,
        block_size_y=dy,
    )

    mesh2d_input = self.meshkernel  # mk.MeshKernel()
    mesh2d_input.curvilinear_compute_rectangular_grid(params)
    mesh2d_input.curvilinear_convert_to_mesh2d()  # convert to ugrid/mesh2d

create_triangular(geometry_list)

Create triangular grid within GeometryList object

Parameters:

Name	Type	Description	Default
geometry_list	GeometryList	GeometryList represeting a polygon within which the mesh is generated.	required

Source code in hydrolib/core/dflowfm/net/models.py

def create_triangular(self, geometry_list: mk.GeometryList) -> None:
    """Create triangular grid within GeometryList object

    Args:
        geometry_list (mk.GeometryList): GeometryList represeting a polygon within which the mesh is generated.
    """
    # Call meshkernel
    self.meshkernel.mesh2d_make_triangular_mesh_from_polygon(geometry_list)

get_mesh2d()

Get the mesh2d as represented in the MeshKernel

Returns:

Type	Description
Mesh2d	The mesh2d as represented in the MeshKernel

Source code in hydrolib/core/dflowfm/net/models.py

def get_mesh2d(self) -> mk.Mesh2d:
    """Get the mesh2d as represented in the MeshKernel

    Returns:
        (mk.Mesh2d): The mesh2d as represented in the MeshKernel
    """
    return self.meshkernel.mesh2d_get()

is_empty()

Determine whether this Mesh2d is empty.

Returns:

Type	Description
bool	Whether this Mesh2d is empty.

Source code in hydrolib/core/dflowfm/net/models.py

def is_empty(self) -> bool:
    """Determine whether this Mesh2d is empty.

    Returns:
        (bool): Whether this Mesh2d is empty.
    """
    return self.mesh2d_node_x.size == 0

read_file(file_path)

Read the Mesh2d from the file at file_path.

Parameters:

Name	Type	Description	Default
file_path	Path	Path to the file to be read.	required

Source code in hydrolib/core/dflowfm/net/models.py

def read_file(self, file_path: Path) -> None:
    """Read the Mesh2d from the file at file_path.

    Args:
        file_path (Path): Path to the file to be read.
    """
    reader = UgridReader(file_path)
    reader.read_mesh2d(self)

refine(polygon, level, parameters=None)

Refine the mesh within a polygon, by a number of steps (level)

Parameters:

Name	Type	Description	Default
polygon	GeometryList	Polygon in which to refine	required
level	int	Number of refinement steps	required
parameters	MeshRefinementParameters	object containing the Meshkernel refinement parameters. Default values are taken from meskernel documentation.	None

Source code in hydrolib/core/dflowfm/net/models.py

def refine(
    self,
    polygon: mk.GeometryList,
    level: int,
    parameters: mk.MeshRefinementParameters = None,
):
    """Refine the mesh within a polygon, by a number of steps (level)

    Args:
        polygon (GeometryList): Polygon in which to refine
        level (int): Number of refinement steps
        parameters (MeshRefinementParameters): object containing the Meshkernel refinement parameters. Default values are taken from meskernel documentation.
    """
    if parameters is None:
        parameters = mk.MeshRefinementParameters(
            refine_intersected=True,
            use_mass_center_when_refining=True,
            min_edge_size=1.0,
            refinement_type=2,
            connect_hanging_nodes=True,
            account_for_samples_outside_face=True,
            max_refinement_iterations=1,
            smoothing_iterations=5,
            max_courant_time=120.0,
            directional_refinement=False,
        )
    parameters.max_refinement_iterations = level
    self.meshkernel.mesh2d_refine_based_on_polygon(polygon, parameters)

Network

Source code in hydrolib/core/dflowfm/net/models.py

class Network:
    def __init__(self, is_geographic: bool = False) -> None:
        if not is_geographic:
            projection = mk.ProjectionType.CARTESIAN
        else:
            projection = mk.ProjectionType.SPHERICAL

        self.meshkernel = mk.MeshKernel(projection=projection)
        self._mesh1d = Mesh1d(meshkernel=self.meshkernel)
        self._mesh2d = Mesh2d(meshkernel=self.meshkernel)
        self._link1d2d = Link1d2d(meshkernel=self.meshkernel)

        # Spatial index (rtree)
        # self._idx = index.Index()

    @classmethod
    def from_file(cls, file_path: Path) -> Network:
        """Read network from file. This classmethod checks what mesh components (mesh1d & network1d, mesh2d, link1d2d) are
        present, and loads them one by one.

        Args:
            file_path (Path): path to netcdf file with network data

        Returns:
            Network: The instance of the class itself that is returned
        """

        network = cls()
        ds = nc.Dataset(file_path)  # type: ignore[import]

        reader = UgridReader(file_path)

        reader.read_mesh1d_network1d(network._mesh1d)
        reader.read_mesh2d(network._mesh2d)
        reader.read_link1d2d(network._link1d2d)

        ds.close()

        return network

    def to_file(self, file: Path) -> None:
        """Write network to file

        Args:
            file (Path): File where _net.nc is written to.
        """

        writer = UgridWriter()
        writer.write(self, file)

    @property
    def is_geographic(self) -> bool:
        """Whether or not this network has a geographic projection.

        Returns:
            bool: True if this network is geographic; otherwise, False.
        """
        projection = self.meshkernel.get_projection()
        if projection == mk.ProjectionType.CARTESIAN:
            return False
        else:
            return True

    def link1d2d_from_1d_to_2d(
        self, branchids: List[str] = None, polygon: GeometryList = None
    ) -> None:
        self._mesh1d._set_mesh1d()

        node_mask = self._mesh1d.get_node_mask(branchids)
        if polygon is None:
            polygon = self.meshkernel.mesh2d_get_mesh_boundaries_as_polygons()

        self._link1d2d._link_from_1d_to_2d(node_mask, polygon=polygon)

    def mesh2d_create_rectilinear_within_extent(
        self, extent: tuple, dx: float, dy: float
    ) -> None:
        self._mesh2d.create_rectilinear(extent=extent, dx=dx, dy=dy)

    def mesh2d_create_triangular_within_polygon(self, polygon: mk.GeometryList) -> None:
        """Create triangular grid within GeometryList object. Calls _mesh2d.create_triangular
        directly, but is easier accessible for users.

        Args:
            polygon (mk.GeometryList): GeometryList representing a polygon within which the mesh is generated.
        """
        self._mesh2d.create_triangular(geometry_list=polygon)

    def mesh2d_clip_mesh(
        self,
        geometrylist: mk.GeometryList,
        deletemeshoption: mk.DeleteMeshOption = mk.DeleteMeshOption.INSIDE_NOT_INTERSECTED,
        inside=True,
    ) -> None:
        self._mesh2d.clip(
            geometrylist=geometrylist,
            deletemeshoption=deletemeshoption,
            inside=inside,
        )

    def mesh2d_refine_mesh(
        self,
        polygon: mk.GeometryList,
        level: int = None,
        parameters: mk.MeshRefineParameters = None,
    ):
        self._mesh2d.refine(polygon=polygon, level=level, parameters=parameters)

    def mesh1d_add_branch(
        self,
        branch: Branch,
        name: str = None,
        branch_order: int = -1,
        long_name: str = None,
        force_midpoint: bool = True,
    ) -> None:
        name = self._mesh1d._add_branch(
            branch=branch,
            name=name,
            branch_order=branch_order,
            long_name=long_name,
            force_midpoint=force_midpoint,
        )
        self._mesh1d._set_mesh1d()
        return name

    def plot(self, ax=None):
        """Create a plot of the 1d2d links and edges within this network.

        Args:
            ax (matplotlib.pyplot.Axes, optional): The axes where to plot the edges. Defaults to None.
        """
        import matplotlib.pyplot as plt

        if ax is None:
            _, ax = plt.subplots()
        mesh2d_output = self._mesh2d.get_mesh2d()
        mesh1d_output = self._mesh1d._get_mesh1d()
        links_output = self._link1d2d.meshkernel.contacts_get()
        mesh2d_output.plot_edges(ax=ax, color="r")
        mesh1d_output.plot_edges(ax=ax, color="g")
        links_output.plot_edges(
            ax=ax, mesh1d=mesh1d_output, mesh2d=mesh2d_output, color="k"
        )

is_geographic property

Whether or not this network has a geographic projection.

Returns:

Name	Type	Description
bool	bool	True if this network is geographic; otherwise, False.

from_file(file_path) classmethod

Read network from file. This classmethod checks what mesh components (mesh1d & network1d, mesh2d, link1d2d) are present, and loads them one by one.

Parameters:

Name	Type	Description	Default
file_path	Path	path to netcdf file with network data	required

Returns:

Name	Type	Description
Network	Network	The instance of the class itself that is returned

Source code in hydrolib/core/dflowfm/net/models.py

@classmethod
def from_file(cls, file_path: Path) -> Network:
    """Read network from file. This classmethod checks what mesh components (mesh1d & network1d, mesh2d, link1d2d) are
    present, and loads them one by one.

    Args:
        file_path (Path): path to netcdf file with network data

    Returns:
        Network: The instance of the class itself that is returned
    """

    network = cls()
    ds = nc.Dataset(file_path)  # type: ignore[import]

    reader = UgridReader(file_path)

    reader.read_mesh1d_network1d(network._mesh1d)
    reader.read_mesh2d(network._mesh2d)
    reader.read_link1d2d(network._link1d2d)

    ds.close()

    return network

mesh2d_create_triangular_within_polygon(polygon)

Create triangular grid within GeometryList object. Calls _mesh2d.create_triangular directly, but is easier accessible for users.

Parameters:

Name	Type	Description	Default
polygon	GeometryList	GeometryList representing a polygon within which the mesh is generated.	required

Source code in hydrolib/core/dflowfm/net/models.py

def mesh2d_create_triangular_within_polygon(self, polygon: mk.GeometryList) -> None:
    """Create triangular grid within GeometryList object. Calls _mesh2d.create_triangular
    directly, but is easier accessible for users.

    Args:
        polygon (mk.GeometryList): GeometryList representing a polygon within which the mesh is generated.
    """
    self._mesh2d.create_triangular(geometry_list=polygon)

plot(ax=None)

Create a plot of the 1d2d links and edges within this network.

Parameters:

Name	Type	Description	Default
ax	Axes	The axes where to plot the edges. Defaults to None.	None

Source code in hydrolib/core/dflowfm/net/models.py

def plot(self, ax=None):
    """Create a plot of the 1d2d links and edges within this network.

    Args:
        ax (matplotlib.pyplot.Axes, optional): The axes where to plot the edges. Defaults to None.
    """
    import matplotlib.pyplot as plt

    if ax is None:
        _, ax = plt.subplots()
    mesh2d_output = self._mesh2d.get_mesh2d()
    mesh1d_output = self._mesh1d._get_mesh1d()
    links_output = self._link1d2d.meshkernel.contacts_get()
    mesh2d_output.plot_edges(ax=ax, color="r")
    mesh1d_output.plot_edges(ax=ax, color="g")
    links_output.plot_edges(
        ax=ax, mesh1d=mesh1d_output, mesh2d=mesh2d_output, color="k"
    )

to_file(file)

Write network to file

Parameters:

Name	Type	Description	Default
file	Path	File where _net.nc is written to.	required

Source code in hydrolib/core/dflowfm/net/models.py

def to_file(self, file: Path) -> None:
    """Write network to file

    Args:
        file (Path): File where _net.nc is written to.
    """

    writer = UgridWriter()
    writer.write(self, file)

NetworkModel

Bases: ParsableFileModel

Network model representation.

Source code in hydrolib/core/dflowfm/net/models.py

class NetworkModel(ParsableFileModel):
    """Network model representation."""

    network: Network = Field(default_factory=Network)

    def _post_init_load(self) -> None:
        """
        Load the network file if the filepath exists relative to the
        current FileLoadContext.
        """
        super()._post_init_load()

        if self.filepath is None:
            return

        with file_load_context() as context:
            network_path = context.resolve(self.filepath)

            if network_path.is_file():
                self.network = Network.from_file(network_path)

    @property
    def _mesh1d(self):
        return self.network._mesh1d

    @property
    def _mesh2d(self):
        return self.network._mesh2d

    @property
    def _link1d2d(self):
        return self.network._link1d2d

    @classmethod
    def _ext(cls) -> str:
        return ".nc"

    @classmethod
    def _filename(cls) -> str:
        return "network"

    def _save(self, save_settings: ModelSaveSettings):
        with file_load_context() as context:
            write_path = context.resolve(self.filepath)  # type: ignore[arg-type]

            write_path.parent.mkdir(parents=True, exist_ok=True)
            self.network.to_file(write_path)

    def _export(self, folder: Path) -> None:
        filename = Path(self.filepath.name) if self.filepath else self._generate_name()
        self.filepath = folder / filename
        folder.mkdir(parents=True, exist_ok=True)
        self.network.to_file(self.filepath)

    def _parse(self, _):
        return {}

    @classmethod
    def _get_serializer(cls):
        # Unused, but requires abstract implementation
        pass

    @classmethod
    def _get_parser(cls):
        # Unused, but requires abstract implementation
        pass

    @property
    def plot(self):
        return self.network.plot

split_by(gl, by)

Function to split mk.GeometryList by seperator.

Parameters:

Name	Type	Description	Default
gl	GeometryList	The geometry list to split.	required
by	float	The value by which to split the gl.	required

Returns:

Name	Type	Description
list	list	The split lists.

Source code in hydrolib/core/dflowfm/net/models.py

def split_by(gl: mk.GeometryList, by: float) -> list:
    """Function to split mk.GeometryList by seperator.

    Args:
        gl (mk.GeometryList): The geometry list to split.
        by (float): The value by which to split the gl.

    Returns:
        list: The split lists.
    """
    x, y = gl.x_coordinates.copy(), gl.y_coordinates.copy()
    idx = np.where(x == by)[0]

    xparts = np.split(x, idx)
    yparts = np.split(y, idx)

    lists = [
        mk.GeometryList(xp[min(i, 1) :], yp[min(i, 1) :])
        for i, (xp, yp) in enumerate(zip(xparts, yparts))
    ]

    return lists