Skip to content

lonboard.traits

lonboard.traits.ColorAccessor

Bases: FixedErrorTraitType

A trait to validate input for a deck.gl color accessor.

Various input is allowed:

  • A list or tuple with three or four integers, ranging between 0 and 255 (inclusive). This will be used as the color for all objects.
  • A str representing a hex color or "well known" color interpretable by matplotlib.colors.to_rgba. See also matplotlib's list of named colors.
  • A numpy ndarray with two dimensions and data type np.uint8. The size of the second dimension must be 3 or 4, and will correspond to either RGB or RGBA colors.
  • A pyarrow FixedSizeListArray or ChunkedArray containing FixedSizeListArrays. The inner size of the fixed size list must be 3 or 4 and its child must have type uint8.
  • Any Arrow fixed size list array from a library that implements the Arrow PyCapsule Interface.

You can use helpers in the lonboard.colormap module (i.e. apply_continuous_cmap) to simplify constructing numpy arrays for color values.

lonboard.traits.DashArrayAccessor

Bases: FixedErrorTraitType

A trait to validate input for a deck.gl dash accessor.

Primarily used in PathStyleExtension.

Various input is allowed:

  • A list or tuple with 2 integers. This defines the dash size and gap size respectively.
  • A numpy ndarray with two dimensions and numeric data type. The size of the second dimension must be 2.
  • A pyarrow FixedSizeListArray or ChunkedArray containing FixedSizeListArrays. The inner size of the fixed size list must be 2.
  • Any Arrow fixed size list array from a library that implements the Arrow PyCapsule Interface.

lonboard.traits.FloatAccessor

Bases: FixedErrorTraitType

A trait to validate input for a deck.gl float accessor.

Various input is allowed:

  • An int or float. This will be used as the value for all objects.
  • A numpy ndarray with a numeric data type. This will be casted to an array of data type np.float32. Each value in the array will be used as the value for the object at the same row index.
  • A pandas Series with a numeric data type. This will be casted to an array of data type np.float32. Each value in the array will be used as the value for the object at the same row index.
  • A pyarrow FloatArray, DoubleArray or ChunkedArray containing either a FloatArray or DoubleArray. Each value in the array will be used as the value for the object at the same row index.
  • Any Arrow floating point array from a library that implements the Arrow PyCapsule Interface. This includes data structures from geoarrow-rust.

lonboard.traits.FilterValueAccessor

Bases: FixedErrorTraitType

Validate input for get_filter_value.

A trait to validate input for the get_filter_value accessor added by the DataFilterExtension, which can have between 1 and 4 float values per row.

Various input is allowed:

  • An int or float. This will be used as the value for all objects. The filter_size of the DataFilterExtension instance must be 1.
  • A one-dimensional numpy ndarray with a numeric data type. This will be casted to an array of data type np.float32. Each value in the array will be used as the value for the object at the same row index. The filter_size of the DataFilterExtension instance must be 1.
  • A two-dimensional numpy ndarray with a numeric data type. This will be casted to an array of data type np.float32. Each value in the array will be used as the value for the object at the same row index. The filter_size of the DataFilterExtension instance must match the size of the second dimension of the array.
  • A pandas Series with a numeric data type. This will be casted to an array of data type np.float32. Each value in the array will be used as the value for the object at the same row index. The filter_size of the DataFilterExtension instance must be 1.
  • A pyarrow FloatArray, DoubleArray or ChunkedArray containing either a FloatArray or DoubleArray. Each value in the array will be used as the value for the object at the same row index. The filter_size of the DataFilterExtension instance must be 1.

Alternatively, you can pass any corresponding Arrow data structure from a library that implements the Arrow PyCapsule Interface. - A pyarrow FixedSizeListArray or ChunkedArray containing FixedSizeListArrays. The child array of the fixed size list must be of floating point type. The filter_size of the DataFilterExtension instance must match the list size.

Alternatively, you can pass any corresponding Arrow data structure from a library that implements the Arrow PyCapsule Interface.

lonboard.traits.NormalAccessor

Bases: FixedErrorTraitType

A representation of a deck.gl "normal" accessor.

This is primarily used in the [lonboard.PointCloudLayer].

Acceptable inputs: - A list or tuple with three int or float values. This will be used as the normal for all objects. - A numpy ndarray with two dimensions and floating point type. The size of the second dimension must be 3, i.e. its shape must be (N, 3). - a pyarrow FixedSizeListArray or ChunkedArray containing FixedSizeListArrays where the size of the inner fixed size list 3. The child array must have type float32. - Any Arrow array that matches the above restrictions from a library that implements the Arrow PyCapsule Interface.

lonboard.traits.PointAccessor

Bases: FixedErrorTraitType

A representation of a deck.gl point accessor.

Various input is allowed:

  • A numpy ndarray with two dimensions and data type np.float64. The size of the second dimension must be 2 or 3, and will correspond to either XY or XYZ positions.
  • A pyarrow FixedSizeListArray, StructArray or ChunkedArray containing FixedSizeListArrays or StructArrays. The inner size of the fixed size list must be 2 or 3 and its child must be of floating point type.
  • Any Arrow fixed size list or struct array from a library that implements the Arrow PyCapsule Interface.

lonboard.traits.ArrowTableTrait

Bases: FixedErrorTraitType

A trait to validate input for a geospatial Arrow-backed table.

Allowed input includes:

lonboard.experimental.traits.TimestampAccessor

Bases: FixedErrorTraitType

A representation of a deck.gl coordinate-timestamp accessor.

deck.gl handles timestamps on the GPU as float32 values. This class will validate that the input timestamps are representable as float32 integers, and will automatically reduce the precision of input data if necessary to fit inside a float32.

Accepted input includes:

  • A pyarrow ListArray containing a temporal array such as a TimestampArray. Each value in the array will be used as the value for the object at the same row index.
  • Any Arrow list array containing a temporal array from a library that implements the Arrow PyCapsule Interface.