Datasets

class Dataset

Dynamic homogeneous multi-dimensional numerical data stored in std::vector. Used to record data collected during an ExperimentalRun.

navground.sim.Dataset supports the buffer protocol and is convertible to and from a numpy.ndarray:

>>> from navground import sim
>>> import numpy
>>>
>>> ds = sim.Dataset(data=numpy.array([1, 2, 3], dtype=numpy.uint8))
>>> ds
<Dataset: shape (3,), dtype uint8>
>>> numpy.asarray(ds)
array([1, 2, 3], dtype=uint8)

__init__(data: object = None, dtype: object = None, item_shape: list[int] = []) → None

Instantiate a dataset with an object convertible to a numpy array of one of the types of navground::sim::Dataset::Scalar.

Parameters:

• data (numpy.typing.ArrayLike | None) – Copies shape, data and dtype from this numpy array (if specified)

• dtype (numpy.typing.DTypeLike | None.) – The type of data to store (if specified and if data is None)

• item_shape (list[int]) – Set the shape of all axis except the first (only if data is None ). Set to an empty list to instantiate a flat dataset.

append(values: object, reset: bool = False) → None

Append items from (objects convertible to) numpy arrays of one of the types of navground::sim::Dataset::Scalar.

Parameters:

• values (numpy.typing.ArrayLike) – Append data and dtype from this numpy array

• reset (bool) – Whether to replace the data instead of appending.

get_buffer(index: int) → navground.core.Buffer

Copy one dataset entry to a new buffer

Parameters:

index – The index of the entry. Set to -1 to add all entries.

Returns:

The buffer

push(value: object) → None

Add an item of one of the types of navground::sim::Dataset::Scalar.

The item will be converted to the current data type, see dtype().

Parameters:

value (numbers.Real) – The value to add

reset() → None

Clear all data.

write_buffer(buffer: navground.core.Buffer, index: int) → bool

Writes a buffer with the content of one dataset entry.

Parameters:

• buffer – The buffer

• index – The index of the entry

Returns:

True if writing was successful.

property dtype

Type:

numpy.dtype

The type of the dataset.

Can be set to any object that is convertible to numpy.dtype.

property is_valid

Type:

bool

Determines if valid.

property item_shape

Type:

list[int]

The shape of all axis except the first.

property shape

Type:

list[int]

Returns the shape of the multi-dimensional dataset