Datatypes

`rerun.datatypes`

`class Angle`

Bases: AngleExt

Datatype: Angle in either radians or degrees.

`inner: float = field(converter=float)` `class-attribute` `instance-attribute`

Must be one of:

Radians (float): Angle in radians. One turn is equal to 2π (or τ) radians. Only one of degrees or radians should be set.
Degrees (float): Angle in degrees. One turn is equal to 360 degrees. Only one of degrees or radians should be set.

`kind: Literal['radians', 'degrees'] = field(default='radians')` `class-attribute` `instance-attribute`

Possible values:

"radians": Angle in radians. One turn is equal to 2π (or τ) radians. Only one of degrees or radians should be set.
"degrees": Angle in degrees. One turn is equal to 360 degrees. Only one of degrees or radians should be set.

`def init(rad=None, deg=None)`

Create a new instance of the Angle datatype.

PARAMETER	DESCRIPTION
`rad`	Angle in radians, specify either `rad` or `deg`. TYPE: `float \| None` DEFAULT: `None`
`deg`	Angle in degrees, specify either `rad` or `deg`. TYPE: `float \| None` DEFAULT: `None`

`class AnnotationInfo`

Bases: AnnotationInfoExt

Datatype: Annotation info annotating a class id or key-point id.

Color and label will be used to annotate entities/keypoints which reference the id. The id refers either to a class or key-point id

`def init(id, label=None, color=None)`

Create a new instance of the AnnotationInfo datatype.

PARAMETER	DESCRIPTION
`id`	`ClassId` or `KeypointId` to which this annotation info belongs. TYPE: `int`
`label`	The label that will be shown in the UI. TYPE: `Utf8Like \| None` DEFAULT: `None`
`color`	The color that will be applied to the annotated entity. TYPE: `Rgba32Like \| None` DEFAULT: `None`

`class Bool`

Datatype: A single boolean.

`def init(value)`

Create a new instance of the Bool datatype.

`class ClassDescription`

Bases: ClassDescriptionExt

Datatype: The description of a semantic Class.

If an entity is annotated with a corresponding ClassId, Rerun will use the attached AnnotationInfo to derive labels and colors.

Keypoints within an annotation class can similarly be annotated with a KeypointId in which case we should defer to the label and color for the AnnotationInfo specifically associated with the Keypoint.

Keypoints within the class can also be decorated with skeletal edges. Keypoint-connections are pairs of KeypointIds. If an edge is defined, and both keypoints exist within the instance of the class, then the keypoints should be connected with an edge. The edge should be labeled and colored as described by the class's AnnotationInfo.

Note that a ClassDescription can be directly logged using rerun.log. This is equivalent to logging a rerun.AnnotationContext containing a single ClassDescription.

`def init(*, info, keypoint_annotations=[], keypoint_connections=[])`

Create a new instance of the ClassDescription datatype.

PARAMETER	DESCRIPTION
`info`	The `AnnotationInfo` for the class. TYPE: `AnnotationInfoLike`
`keypoint_annotations`	The `AnnotationInfo` for all the keypoints. TYPE: `Sequence[AnnotationInfoLike] \| None` DEFAULT: `[]`
`keypoint_connections`	The connections between keypoints. TYPE: `Sequence[KeypointPairLike] \| None` DEFAULT: `[]`

`class ClassDescriptionMapElem`

Bases: ClassDescriptionMapElemExt

Datatype: A helper type for mapping class IDs to class descriptions.

This is internal to the AnnotationContext structure.

`def init(class_id, class_description)`

Create a new instance of the ClassDescriptionMapElem datatype.

PARAMETER	DESCRIPTION
`class_id`	The key: the class ID. TYPE: `ClassIdLike`
`class_description`	The value: class name, color, etc. TYPE: `ClassDescriptionLike`

`class ClassId`

Datatype: A 16-bit ID representing a type of semantic class.

`def init(id)`

Create a new instance of the ClassId datatype.

`class EntityPath`

Datatype: A path to an entity in the DataStore.

`def init(path)`

Create a new instance of the EntityPath datatype.

`class Float32`

Datatype: A single-precision 32-bit IEEE 754 floating point number.

`def init(value)`

Create a new instance of the Float32 datatype.

`class KeypointId`

Datatype: A 16-bit ID representing a type of semantic keypoint within a class.

KeypointIds are only meaningful within the context of a [rerun.datatypes.ClassDescription].

Used to look up an [rerun.datatypes.AnnotationInfo] for a Keypoint within the [rerun.components.AnnotationContext].

`def init(id)`

Create a new instance of the KeypointId datatype.

`class KeypointPair`

Bases: KeypointPairExt

Datatype: A connection between two Keypoints.

`def init(keypoint0, keypoint1)`

Create a new instance of the KeypointPair datatype.

PARAMETER	DESCRIPTION
`keypoint0`	The first point of the pair. TYPE: `KeypointIdLike`
`keypoint1`	The second point of the pair. TYPE: `KeypointIdLike`

`class Mat3x3`

Bases: Mat3x3Ext

Datatype: A 3x3 Matrix.

Matrices in Rerun are stored as flat list of coefficients in column-major order:

            column 0       column 1       column 2
       -------------------------------------------------
row 0 | flat_columns[0] flat_columns[3] flat_columns[6]
row 1 | flat_columns[1] flat_columns[4] flat_columns[7]
row 2 | flat_columns[2] flat_columns[5] flat_columns[8]

However, construction is done from a list of rows, which follows NumPy's convention:

np.testing.assert_array_equal(
    rr.datatypes.Mat3x3([1, 2, 3, 4, 5, 6, 7, 8, 9]).flat_columns, np.array([1, 4, 7, 2, 5, 8, 3, 6, 9], dtype=np.float32)
)
np.testing.assert_array_equal(
    rr.datatypes.Mat3x3([[1, 2, 3], [4, 5, 6], [7, 8, 9]]).flat_columns,
    np.array([1, 4, 7, 2, 5, 8, 3, 6, 9], dtype=np.float32),
)

If you want to construct a matrix from a list of columns instead, use the named columns parameter:

np.testing.assert_array_equal(
    rr.datatypes.Mat3x3(columns=[1, 2, 3, 4, 5, 6, 7, 8, 9]).flat_columns,
    np.array([1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=np.float32),
)
np.testing.assert_array_equal(
    rr.datatypes.Mat3x3(columns=[[1, 2, 3], [4, 5, 6], [7, 8, 9]]).flat_columns,
    np.array([1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=np.float32),
)

`def init(rows=None, *, columns=None)`

`class Mat4x4`

Bases: Mat4x4Ext

Datatype: A 4x4 Matrix.

Matrices in Rerun are stored as flat list of coefficients in column-major order:

           column 0         column 1         column 2         column 3
       --------------------------------------------------------------------
row 0 | flat_columns[0]  flat_columns[4]  flat_columns[8]  flat_columns[12]
row 1 | flat_columns[1]  flat_columns[5]  flat_columns[9]  flat_columns[13]
row 2 | flat_columns[2]  flat_columns[6]  flat_columns[10] flat_columns[14]
row 3 | flat_columns[3]  flat_columns[7]  flat_columns[11] flat_columns[15]

However, construction is done from a list of rows, which follows NumPy's convention:

np.testing.assert_array_equal(
    rr.datatypes.Mat4x4([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]).flat_columns,
    np.array([1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15, 4, 8, 12, 16], dtype=np.float32),
)
np.testing.assert_array_equal(
    rr.datatypes.Mat4x4([[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16]]).flat_columns,
    np.array([1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15, 4, 8, 12, 16], dtype=np.float32),
)

If you want to construct a matrix from a list of columns instead, use the named columns parameter:

np.testing.assert_array_equal(
    rr.datatypes.Mat4x4(columns=[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]).flat_columns,
    np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype=np.float32),
)
np.testing.assert_array_equal(
    rr.datatypes.Mat4x4(columns=[[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16]]).flat_columns,
    np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], dtype=np.float32),
)

`def init(rows=None, *, columns=None)`

`class Material`

Bases: MaterialExt

Datatype: Material properties of a mesh, e.g. its color multiplier.

`def init(albedo_factor=None)`

Create a new instance of the Material datatype.

PARAMETER	DESCRIPTION
`albedo_factor`	Optional color multiplier. TYPE: `Rgba32Like \| None` DEFAULT: `None`

`class Quaternion`

Bases: QuaternionExt

Datatype: A Quaternion represented by 4 real numbers.

Note: although the x,y,z,w components of the quaternion will be passed through to the datastore as provided, when used in the Viewer Quaternions will always be normalized.

`class Range1D`

Bases: Range1DExt

Datatype: A 1D range, specifying a lower and upper bound.

`def init(range)`

Create a new instance of the Range1D datatype.

`class Range2D`

Datatype: An Axis-Aligned Bounding Box in 2D space, implemented as the minimum and maximum corners.

`def init(x_range, y_range)`

Create a new instance of the Range2D datatype.

PARAMETER	DESCRIPTION
`x_range`	The range of the X-axis (usually left and right bounds). TYPE: `Range1DLike`
`y_range`	The range of the Y-axis (usually top and bottom bounds). TYPE: `Range1DLike`

`class Rgba32`

Bases: Rgba32Ext

Datatype: An RGBA color with unmultiplied/separate alpha, in sRGB gamma space with linear alpha.

The color is stored as a 32-bit integer, where the most significant byte is R and the least significant byte is A.

Float colors are assumed to be in 0-1 gamma sRGB space. All other colors are assumed to be in 0-255 gamma sRGB space. If there is an alpha, we assume it is in linear space, and separate (NOT pre-multiplied).

`def init(rgba)`

Create a new instance of the Rgba32 datatype.

`class Rotation3D`

Bases: Rotation3DExt

Datatype: A 3D rotation.

`inner: Union[datatypes.Quaternion, datatypes.RotationAxisAngle] = field(converter=Rotation3DExt.inner__field_converter_override)` `class-attribute` `instance-attribute`

Must be one of:

Quaternion (datatypes.Quaternion): Rotation defined by a quaternion.
AxisAngle (datatypes.RotationAxisAngle): Rotation defined with an axis and an angle.

`class RotationAxisAngle`

Bases: RotationAxisAngleExt

Datatype: 3D rotation represented by a rotation around a given axis.

`def init(axis, angle=None, *, radians=None, degrees=None)`

Create a new instance of the RotationAxisAngle datatype.

PARAMETER	DESCRIPTION
`axis`	Axis to rotate around. This is not required to be normalized. If normalization fails (typically because the vector is length zero), the rotation is silently ignored. TYPE: `Vec3DLike`
`angle`	How much to rotate around the axis. TYPE: `AngleLike \| None` DEFAULT: `None`
`radians`	How much to rotate around the axis, in radians. Specify this instead of `degrees` or `angle`. TYPE: `float \| None` DEFAULT: `None`
`degrees`	How much to rotate around the axis, in radians. Specify this instead of `radians` or `angle`. TYPE: `float \| None` DEFAULT: `None`

`class Scale3D`

Bases: Scale3DExt

Datatype: 3D scaling factor, part of a transform representation.

Example

# uniform scaling
scale = rr.datatypes.Scale3D(3.)

# non-uniform scaling
scale = rr.datatypes.Scale3D([1, 1, -1])
scale = rr.datatypes.Scale3D(rr.datatypes.Vec3D([1, 1, -1]))

`inner: Union[datatypes.Vec3D, float] = field(converter=Scale3DExt.inner__field_converter_override)` `class-attribute` `instance-attribute`

Must be one of:

ThreeD (datatypes.Vec3D): Individual scaling factors for each axis, distorting the original object.
Uniform (float): Uniform scaling factor along all axis.

`class TensorBuffer`

Bases: TensorBufferExt

Datatype: The underlying storage for a Tensor.

Tensor elements are stored in a contiguous buffer of a single type.

`inner: Union[npt.NDArray[np.float16], npt.NDArray[np.float32], npt.NDArray[np.float64], npt.NDArray[np.int16], npt.NDArray[np.int32], npt.NDArray[np.int64], npt.NDArray[np.int8], npt.NDArray[np.uint16], npt.NDArray[np.uint32], npt.NDArray[np.uint64], npt.NDArray[np.uint8]] = field(converter=TensorBufferExt.inner__field_converter_override)` `class-attribute` `instance-attribute`

Must be one of:

U8 (npt.NDArray[np.uint8]): 8bit unsigned integer.
U16 (npt.NDArray[np.uint16]): 16bit unsigned integer.
U32 (npt.NDArray[np.uint32]): 32bit unsigned integer.
U64 (npt.NDArray[np.uint64]): 64bit unsigned integer.
I8 (npt.NDArray[np.int8]): 8bit signed integer.
I16 (npt.NDArray[np.int16]): 16bit signed integer.
I32 (npt.NDArray[np.int32]): 32bit signed integer.
I64 (npt.NDArray[np.int64]): 64bit signed integer.
F16 (npt.NDArray[np.float16]): 16bit IEEE-754 floating point, also known as half.
F32 (npt.NDArray[np.float32]): 32bit IEEE-754 floating point, also known as float or single.
F64 (npt.NDArray[np.float64]): 64bit IEEE-754 floating point, also known as double.
JPEG (npt.NDArray[np.uint8]): Raw bytes of a JPEG file.
NV12 (npt.NDArray[np.uint8]): NV12 is a YUV 4:2:0 chroma downsamples format with 8 bits per channel.

First comes entire image in Y, followed by interleaved lines ordered as U0, V0, U1, V1, etc.
YUY2 (npt.NDArray[np.uint8]): YUY2, also known as YUYV is a YUV 4:2:2 chroma downsampled format with 8 bits per channel.

The order of the channels is Y0, U0, Y1, V0.

`kind: Literal['u8', 'u16', 'u32', 'u64', 'i8', 'i16', 'i32', 'i64', 'f16', 'f32', 'f64', 'jpeg', 'nv12', 'yuy2'] = field(default='u8')` `class-attribute` `instance-attribute`

Possible values:

"u8": 8bit unsigned integer.
"u16": 16bit unsigned integer.
"u32": 32bit unsigned integer.
"u64": 64bit unsigned integer.
"i8": 8bit signed integer.
"i16": 16bit signed integer.
"i32": 32bit signed integer.
"i64": 64bit signed integer.
"f16": 16bit IEEE-754 floating point, also known as half.
"f32": 32bit IEEE-754 floating point, also known as float or single.
"f64": 64bit IEEE-754 floating point, also known as double.
"jpeg": Raw bytes of a JPEG file.
"nv12": NV12 is a YUV 4:2:0 chroma downsamples format with 8 bits per channel.

First comes entire image in Y, followed by interleaved lines ordered as U0, V0, U1, V1, etc.
"yuy2": YUY2, also known as YUYV is a YUV 4:2:2 chroma downsampled format with 8 bits per channel.

The order of the channels is Y0, U0, Y1, V0.

`class TensorData`

Bases: TensorDataExt

Datatype: An N-dimensional array of numbers.

The number of dimensions and their respective lengths is specified by the shape field. The dimensions are ordered from outermost to innermost. For example, in the common case of a 2D RGB Image, the shape would be [height, width, channel].

These dimensions are combined with an index to look up values from the buffer field, which stores a contiguous array of typed values.

Note that the buffer may be encoded in a compressed format such as jpeg or in a format with downsampled chroma, such as NV12 or YUY2. For file formats, the shape is used as a hint, for chroma downsampled format the shape has to be the shape of the decoded image.

`def init(*, shape=None, buffer=None, array=None, dim_names=None)`

Construct a TensorData object.

The TensorData object is internally represented by three fields: shape and buffer.

This constructor provides additional arguments 'array', and 'dim_names'. When passing in a multi-dimensional array such as a np.ndarray, the shape and buffer fields will be populated automagically.

PARAMETER	DESCRIPTION
`self`	The TensorData object to construct. TYPE: `Any`
`shape`	The shape of the tensor. If None, and an array is provided, the shape will be inferred from the shape of the array. TYPE: `Sequence[TensorDimensionLike] \| None` DEFAULT: `None`
`buffer`	The buffer of the tensor. If None, and an array is provided, the buffer will be generated from the array. TYPE: `TensorBufferLike \| None` DEFAULT: `None`
`array`	A numpy array (or The array of the tensor. If None, the array will be inferred from the buffer. TYPE: `TensorLike \| None` DEFAULT: `None`
`dim_names`	The names of the tensor dimensions when generating the shape from an array. TYPE: `Sequence[str \| None] \| None` DEFAULT: `None`

`class TensorDimension`

Datatype: A single dimension within a multi-dimensional tensor.

`def init(size, name=None)`

Create a new instance of the TensorDimension datatype.

PARAMETER	DESCRIPTION
`size`	The length of this dimension. TYPE: `int`
`name`	The name of this dimension, e.g. "width", "height", "channel", "batch', …. TYPE: `str \| None` DEFAULT: `None`

`class TensorDimensionIndexSelection`

Datatype: Indexing a specific tensor dimension.

Selecting dimension=2 and index=42 is similar to doing tensor[:, :, 42, :, :, …] in numpy.

`def init(dimension, index)`

Create a new instance of the TensorDimensionIndexSelection datatype.

PARAMETER	DESCRIPTION
`dimension`	The dimension number to select. TYPE: `int`
`index`	The index along the dimension to use. TYPE: `int`

`class TensorDimensionSelection`

Bases: TensorDimensionSelectionExt

Datatype: Selection of a single tensor dimension.

`def init(dimension, *, invert=False)`

Create a new instance of the TensorDimensionSelection datatype.

PARAMETER	DESCRIPTION
`dimension`	The dimension number to select. TYPE: `int`
`invert`	Invert the direction of the dimension. TYPE: `bool` DEFAULT: `False`

`class TimeInt`

Bases: TimeIntExt

Datatype: A 64-bit number describing either nanoseconds OR sequence numbers.

`def init(*, seq=None, seconds=None, nanos=None)`

Create a new instance of the TimeInt datatype.

PARAMETER	DESCRIPTION
`seq`	Time as a sequence number. Mutually exclusive with seconds and nanos. TYPE: `int \| None` DEFAULT: `None`
`seconds`	Time in seconds. Mutually exclusive with seq and nanos. TYPE: `float \| None` DEFAULT: `None`
`nanos`	Time in nanoseconds. Mutually exclusive with seq and seconds. TYPE: `int \| None` DEFAULT: `None`

`class TimeRange`

Datatype: Visible time range bounds for a specific timeline.

`def init(start, end)`

Create a new instance of the TimeRange datatype.

PARAMETER	DESCRIPTION
`start`	Low time boundary for sequence timeline. TYPE: `TimeRangeBoundaryLike`
`end`	High time boundary for sequence timeline. TYPE: `TimeRangeBoundaryLike`

`class TimeRangeBoundary`

Bases: TimeRangeBoundaryExt

Datatype: Left or right boundary of a time range.

`inner: Union[None, datatypes.TimeInt] = field()` `class-attribute` `instance-attribute`

Must be one of:

CursorRelative (datatypes.TimeInt): Boundary is a value relative to the time cursor.
Absolute (datatypes.TimeInt): Boundary is an absolute value.
Infinite (None): The boundary extends to infinity.

`kind: Literal['cursor_relative', 'absolute', 'infinite'] = field(default='cursor_relative')` `class-attribute` `instance-attribute`

Possible values:

"cursor_relative": Boundary is a value relative to the time cursor.
"absolute": Boundary is an absolute value.
"infinite": The boundary extends to infinity.

`def absolute(time=None, *, seq=None, seconds=None, nanos=None)` `staticmethod`

Boundary that is at an absolute time.

PARAMETER	DESCRIPTION
`time`	Absolute time. Mutually exclusive with seq, seconds and nanos. TYPE: `TimeInt \| None` DEFAULT: `None`
`seq`	Absolute time in sequence numbers. Use this for sequence timelines. Mutually exclusive with time, seconds and nanos. TYPE: `int \| None` DEFAULT: `None`
`seconds`	Absolute time in seconds. Use this for time based timelines. Mutually exclusive with time, seq and nanos. TYPE: `float \| None` DEFAULT: `None`
`nanos`	Absolute time in nanoseconds. Use this for time based timelines. Mutually exclusive with time, seq and seconds. TYPE: `int \| None` DEFAULT: `None`

`def cursor_relative(offset=None, *, seq=None, seconds=None, nanos=None)` `staticmethod`

Boundary that is relative to the timeline cursor.

The offset can be positive or negative. An offset of zero (the default) means the cursor time itself.

PARAMETER	DESCRIPTION
`offset`	Offset from the cursor time. Mutually exclusive with seq, seconds and nanos. TYPE: `TimeInt \| None` DEFAULT: `None`
`seq`	Offset in sequence numbers. Use this for sequence timelines. Mutually exclusive with time, seconds and nanos. TYPE: `int \| None` DEFAULT: `None`
`seconds`	Offset in seconds. Use this for time based timelines. Mutually exclusive with time, seq and nanos. TYPE: `float \| None` DEFAULT: `None`
`nanos`	Offset in nanoseconds. Use this for time based timelines. Mutually exclusive with time, seq and seconds. TYPE: `int \| None` DEFAULT: `None`

`def infinite()` `staticmethod`

Boundary that extends to infinity.

Depending on the context, this can mean the beginning or the end of the timeline.

`class Transform3D`

Bases: Transform3DExt

Datatype: Representation of a 3D affine transform.

`inner: Union[datatypes.TranslationAndMat3x3, datatypes.TranslationRotationScale3D] = field()` `class-attribute` `instance-attribute`

Must be one of:

TranslationAndMat3x3 (datatypes.TranslationAndMat3x3): Translation plus a 3x3 matrix for scale, rotation, skew, etc.
TranslationRotationScale (datatypes.TranslationRotationScale3D): Translation, rotation and scale, decomposed.

`class TranslationAndMat3x3`

Bases: TranslationAndMat3x3Ext

Datatype: Representation of an affine transform via a 3x3 affine matrix paired with a translation.

First applies the matrix, then the translation.

`def init(translation=None, mat3x3=None, *, from_parent=False)`

Create a new instance of the TranslationAndMat3x3 datatype.

PARAMETER	DESCRIPTION
`translation`	3D translation, applied after the matrix. TYPE: `Vec3DLike \| None` DEFAULT: `None`
`mat3x3`	3x3 matrix for scale, rotation & shear. TYPE: `Mat3x3Like \| None` DEFAULT: `None`
`from_parent`	If true, the transform maps from the parent space to the space where the transform was logged. Otherwise, the transform maps from the space to its parent. TYPE: `bool` DEFAULT: `False`

`class TranslationRotationScale3D`

Bases: TranslationRotationScale3DExt

Datatype: Representation of an affine transform via separate translation, rotation & scale.

`class UInt32`

Datatype: A 32bit unsigned integer.

`def init(value)`

Create a new instance of the UInt32 datatype.

`class UInt64`

Datatype: A 64bit unsigned integer.

`def init(value)`

Create a new instance of the UInt64 datatype.

`class UVec2D`

Datatype: A uint32 vector in 2D space.

`def init(xy)`

Create a new instance of the UVec2D datatype.

`class UVec3D`

Bases: UVec3DExt

Datatype: A uint32 vector in 3D space.

`def init(xyz)`

Create a new instance of the UVec3D datatype.

`class UVec4D`

Datatype: A uint vector in 4D space.

`def init(xyzw)`

Create a new instance of the UVec4D datatype.

`class Utf8`

Datatype: A string of text, encoded as UTF-8.

`def init(value)`

Create a new instance of the Utf8 datatype.

`class Uuid`

Bases: UuidExt

Datatype: A 16-byte UUID.

`def init(bytes)`

Create a new instance of the Uuid datatype.

PARAMETER	DESCRIPTION
`bytes`	The raw bytes representing the UUID. TYPE: `UuidLike`

`class Vec2D`

Bases: Vec2DExt

Datatype: A vector in 2D space.

`def init(xy)`

Create a new instance of the Vec2D datatype.

`class Vec3D`

Bases: Vec3DExt

Datatype: A vector in 3D space.

`def init(xyz)`

Create a new instance of the Vec3D datatype.

`class Vec4D`

Bases: Vec4DExt

Datatype: A vector in 4D space.

`def init(xyzw)`

Create a new instance of the Vec4D datatype.

`class VisibleTimeRange`

Bases: VisibleTimeRangeExt

Datatype: Visible time range bounds for a specific timeline.

`def init(timeline, range=None, *, start=None, end=None)`

Create a new instance of the VisibleTimeRange datatype.

PARAMETER	DESCRIPTION
`timeline`	Name of the timeline this applies to. TYPE: `Utf8Like`
`range`	Time range to use for this timeline. TYPE: `TimeRangeLike \| None` DEFAULT: `None`
`start`	Low time boundary for sequence timeline. Specify this instead of `range`. TYPE: `TimeRangeBoundary \| None` DEFAULT: `None`
`end`	High time boundary for sequence timeline. Specify this instead of `range`. TYPE: `TimeRangeBoundary \| None` DEFAULT: `None`

Datatypes

rerun.datatypes

class Angle

inner: float = field(converter=float) class-attribute instance-attribute

kind: Literal['radians', 'degrees'] = field(default='radians') class-attribute instance-attribute

def __init__(rad=None, deg=None)

class AnnotationInfo

def __init__(id, label=None, color=None)

class Bool

def __init__(value)

class ClassDescription

def __init__(*, info, keypoint_annotations=[], keypoint_connections=[])

class ClassDescriptionMapElem

def __init__(class_id, class_description)

class ClassId

def __init__(id)

class EntityPath

def __init__(path)

class Float32

def __init__(value)

class KeypointId

def __init__(id)

class KeypointPair

def __init__(keypoint0, keypoint1)

class Mat3x3

def __init__(rows=None, *, columns=None)

class Mat4x4

def __init__(rows=None, *, columns=None)

class Material

def __init__(albedo_factor=None)

class Quaternion

class Range1D

def __init__(range)

class Range2D

def __init__(x_range, y_range)

class Rgba32

def __init__(rgba)

class Rotation3D

inner: Union[datatypes.Quaternion, datatypes.RotationAxisAngle] = field(converter=Rotation3DExt.inner__field_converter_override) class-attribute instance-attribute

class RotationAxisAngle

def __init__(axis, angle=None, *, radians=None, degrees=None)

class Scale3D

inner: Union[datatypes.Vec3D, float] = field(converter=Scale3DExt.inner__field_converter_override) class-attribute instance-attribute

class TensorBuffer

kind: Literal['u8', 'u16', 'u32', 'u64', 'i8', 'i16', 'i32', 'i64', 'f16', 'f32', 'f64', 'jpeg', 'nv12', 'yuy2'] = field(default='u8') class-attribute instance-attribute

class TensorData

def __init__(*, shape=None, buffer=None, array=None, dim_names=None)

class TensorDimension

def __init__(size, name=None)

class TensorDimensionIndexSelection

def __init__(dimension, index)

class TensorDimensionSelection

def __init__(dimension, *, invert=False)

class TimeInt

def __init__(*, seq=None, seconds=None, nanos=None)

class TimeRange

def __init__(start, end)

class TimeRangeBoundary

inner: Union[None, datatypes.TimeInt] = field() class-attribute instance-attribute

kind: Literal['cursor_relative', 'absolute', 'infinite'] = field(default='cursor_relative') class-attribute instance-attribute

def absolute(time=None, *, seq=None, seconds=None, nanos=None) staticmethod

def cursor_relative(offset=None, *, seq=None, seconds=None, nanos=None) staticmethod

def infinite() staticmethod

class Transform3D

inner: Union[datatypes.TranslationAndMat3x3, datatypes.TranslationRotationScale3D] = field() class-attribute instance-attribute

class TranslationAndMat3x3

def __init__(translation=None, mat3x3=None, *, from_parent=False)

class TranslationRotationScale3D

class UInt32

def __init__(value)

class UInt64

def __init__(value)

class UVec2D

def __init__(xy)

class UVec3D

def __init__(xyz)

class UVec4D

def __init__(xyzw)

class Utf8

def __init__(value)

`rerun.datatypes`

`class Angle`

`inner: float = field(converter=float)` `class-attribute` `instance-attribute`

`kind: Literal['radians', 'degrees'] = field(default='radians')` `class-attribute` `instance-attribute`

`def init(rad=None, deg=None)`

`class AnnotationInfo`

`def init(id, label=None, color=None)`

`class Bool`

`def init(value)`

`class ClassDescription`

`def init(*, info, keypoint_annotations=[], keypoint_connections=[])`

`class ClassDescriptionMapElem`

`def init(class_id, class_description)`

`class ClassId`

`def init(id)`

`class EntityPath`

`def init(path)`

`class Float32`

`def init(value)`

`class KeypointId`

`def init(id)`

`class KeypointPair`

`def init(keypoint0, keypoint1)`

`class Mat3x3`

`def init(rows=None, *, columns=None)`

`class Mat4x4`

`def init(rows=None, *, columns=None)`

`class Material`

`def init(albedo_factor=None)`

`class Quaternion`

`class Range1D`

`def init(range)`

`class Range2D`

`def init(x_range, y_range)`

`class Rgba32`

`def init(rgba)`

`class Rotation3D`

`inner: Union[datatypes.Quaternion, datatypes.RotationAxisAngle] = field(converter=Rotation3DExt.inner__field_converter_override)` `class-attribute` `instance-attribute`

`class RotationAxisAngle`

`def init(axis, angle=None, *, radians=None, degrees=None)`

`class Scale3D`

`inner: Union[datatypes.Vec3D, float] = field(converter=Scale3DExt.inner__field_converter_override)` `class-attribute` `instance-attribute`

`class TensorBuffer`

`kind: Literal['u8', 'u16', 'u32', 'u64', 'i8', 'i16', 'i32', 'i64', 'f16', 'f32', 'f64', 'jpeg', 'nv12', 'yuy2'] = field(default='u8')` `class-attribute` `instance-attribute`

`class TensorData`

`def init(*, shape=None, buffer=None, array=None, dim_names=None)`

`class TensorDimension`

`def init(size, name=None)`

`class TensorDimensionIndexSelection`

`def init(dimension, index)`

`class TensorDimensionSelection`

`def init(dimension, *, invert=False)`

`class TimeInt`

`def init(*, seq=None, seconds=None, nanos=None)`

`class TimeRange`

`def init(start, end)`

`class TimeRangeBoundary`

`inner: Union[None, datatypes.TimeInt] = field()` `class-attribute` `instance-attribute`

`kind: Literal['cursor_relative', 'absolute', 'infinite'] = field(default='cursor_relative')` `class-attribute` `instance-attribute`

`def absolute(time=None, *, seq=None, seconds=None, nanos=None)` `staticmethod`

`def cursor_relative(offset=None, *, seq=None, seconds=None, nanos=None)` `staticmethod`

`def infinite()` `staticmethod`

`class Transform3D`

`inner: Union[datatypes.TranslationAndMat3x3, datatypes.TranslationRotationScale3D] = field()` `class-attribute` `instance-attribute`

`class TranslationAndMat3x3`

`def init(translation=None, mat3x3=None, *, from_parent=False)`

`class TranslationRotationScale3D`

`class UInt32`

`def init(value)`

`class UInt64`

`def init(value)`

`class UVec2D`

`def init(xy)`

`class UVec3D`

`def init(xyz)`

`class UVec4D`

`def init(xyzw)`

`class Utf8`

`def init(value)`

`class Uuid`