Transform

This page contains classes defined in hakowan.transform module.

Transform dataclass

Transform is the base class of all transforms.

Source code in hakowan/grammar/transform/transform.py

@dataclass(kw_only=True, slots=True)
class Transform:
    """Transform is the base class of all transforms."""

    _child: Optional["Transform"] = None

    def __imul__(self, other: "Transform") -> "Transform":
        """In place update by applying another transform after the current transform.

        Args:
            other: The transform to apply after the current transform.
        """
        # Because transform may be used in multiple places in the layer graph, and it may have a
        # child in the future, it must be deep copied to avoid undesired side effects.
        if self._child is None:
            self._child = copy.deepcopy(other)
        else:
            t = self._child
            while t._child is not None:
                t = t._child
            t._child = copy.deepcopy(other)
        return self

    def __mul__(self, other: "Transform") -> "Transform":
        """Apply another transform, `other`, after the current transform.

        Args:
            other: The other transform.

        Returns: A new transform that is the composition of the current transform and `other`.
        """
        r = copy.deepcopy(self)
        r *= other
        return r

__imul__(other)

In place update by applying another transform after the current transform.

Parameters:

Name	Type	Description	Default
other	Transform	The transform to apply after the current transform.	required

Source code in hakowan/grammar/transform/transform.py

def __imul__(self, other: "Transform") -> "Transform":
    """In place update by applying another transform after the current transform.

    Args:
        other: The transform to apply after the current transform.
    """
    # Because transform may be used in multiple places in the layer graph, and it may have a
    # child in the future, it must be deep copied to avoid undesired side effects.
    if self._child is None:
        self._child = copy.deepcopy(other)
    else:
        t = self._child
        while t._child is not None:
            t = t._child
        t._child = copy.deepcopy(other)
    return self

__mul__(other)

Apply another transform, other, after the current transform.

Parameters:

Name	Type	Description	Default
other	Transform	The other transform.	required

Returns: A new transform that is the composition of the current transform and other.

Source code in hakowan/grammar/transform/transform.py

def __mul__(self, other: "Transform") -> "Transform":
    """Apply another transform, `other`, after the current transform.

    Args:
        other: The other transform.

    Returns: A new transform that is the composition of the current transform and `other`.
    """
    r = copy.deepcopy(self)
    r *= other
    return r

Filter dataclass

Bases: Transform

Filter data based on a condition.

Attributes:

Name	Type	Description
data	AttributeLike | None	The attribute to filter on. If None, the vertex position is used.
condition	Callable	A callable that takes a single argument, the value of the attribute, and returns a boolean indicating whether the data should be kept.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class Filter(Transform):
    """Filter data based on a condition.

    Attributes:
        data: The attribute to filter on. If None, the vertex position is used.
        condition: A callable that takes a single argument, the value of the attribute, and returns
            a boolean indicating whether the data should be kept.
    """

    data: AttributeLike | None = None
    condition: Callable = lambda x: True

UVMesh dataclass

Bases: Transform

Extract UV mesh from data.

Attributes:

Name	Type	Description
uv	AttributeLike | None	The attribute defining the UV coordinates. If None, automatically deetect the UV attribute from the data.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class UVMesh(Transform):
    """Extract UV mesh from data.

    Attributes:
        uv: The attribute defining the UV coordinates. If None, automatically deetect the UV
            attribute from the data.
    """

    uv: AttributeLike | None = None

Affine dataclass

Bases: Transform

Apply affine transformation to data.

Attributes:

Name	Type	Description
matrix	ArrayLike	The 4x4 affine matrix to apply.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class Affine(Transform):
    """Apply affine transformation to data.

    Attributes:
        matrix: The 4x4 affine matrix to apply.
    """

    matrix: npt.ArrayLike

Compute dataclass

Bases: Transform

Compute new attributes from the current data frame.

Attributes:

Name	Type	Description
x	str | None	Extract the x coordinate as an attribute.
y	str | None	Extract the y coordinate as an attribute.
z	str | None	Extract the z coordinate as an attribute.
normal	str | None	Compute the normal vector field as an attribute.
vertex_normal	str | None	Compute the vertex normal vector field as an attribute.
facet_normal	str | None	Compute the facet normal vector field as an attribute.
component	str | None	Compute connected component ids.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True, kw_only=True)
class Compute(Transform):
    """Compute new attributes from the current data frame.

    Attributes:
        x: Extract the x coordinate as an attribute.
        y: Extract the y coordinate as an attribute.
        z: Extract the z coordinate as an attribute.
        normal: Compute the normal vector field as an attribute.
        vertex_normal: Compute the vertex normal vector field as an attribute.
        facet_normal: Compute the facet normal vector field as an attribute.
        component: Compute connected component ids.
    """

    x: str | None = None
    y: str | None = None
    z: str | None = None
    normal: str | None = None
    vertex_normal: str | None = None
    facet_normal: str | None = None
    component: str | None = None

Explode dataclass

Bases: Transform

Explode data into multiple pieces.

Attributes:

Name	Type	Description
pieces	AttributeLike	The attribute defining the pieces.
magnitude	float	The magnitude of the displacement.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class Explode(Transform):
    """Explode data into multiple pieces.

    Attributes:
        pieces: The attribute defining the pieces.
        magnitude: The magnitude of the displacement.
    """

    pieces: AttributeLike
    magnitude: float = 1

Norm dataclass

Bases: Transform

Compute the row-wise norm of a given vector attribute.

Attributes:

Name	Type	Description
data	AttributeLike	The vector attribute to compute the norm on.
norm_attr_name	str	The name of the output norm attribute.
order	int	The order of the norm. Default is 2, which is the L2 norm.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class Norm(Transform):
    """Compute the row-wise norm of a given vector attribute.

    Attributes:
        data: The vector attribute to compute the norm on.
        norm_attr_name: The name of the output norm attribute.
        order: The order of the norm. Default is 2, which is the L2 norm.
    """

    data: AttributeLike
    norm_attr_name: str
    order: int = 2

Boundary dataclass

Bases: Transform

Compute the boundary of a mesh.

Attributes:

Name	Type	Description
attributes	list[str]	The attributes to take into account when computing the boundary. i.e. discontinuities in these attributes will be considered as boundaries.

Source code in hakowan/grammar/transform/transform.py

@dataclass(slots=True)
class Boundary(Transform):
    """ Compute the boundary of a mesh.

    Attributes:
        attributes: The attributes to take into account when computing the boundary.
            i.e. discontinuities in these attributes will be considered as boundaries.
    """
    attributes: list[str] = field(default_factory=list)