dataeval.utils.arrays.ensure_embeddings¶

dataeval.utils.arrays.ensure_embeddings(embeddings: T, dtype: torch.dtype, unit_interval: True | False | 'force' = False) → torch.Tensor¶

dataeval.utils.arrays.ensure_embeddings(embeddings: T, dtype: type[_np_dtype], unit_interval: True | False | 'force' = False) → numpy.typing.NDArray[_np_dtype]

dataeval.utils.arrays.ensure_embeddings(embeddings: T, dtype: None = None, unit_interval: True | False | 'force' = False) → T

Validates the embeddings array and converts it to the specified type.

Parameters:¶

embeddings : ArrayLike¶

Embeddings array

dtype : numpy dtype or torch dtype or None, default None¶

The desired dtype of the output array, None to skip conversion

unit_interval : bool or "force", default False¶

Whether to validate or force the embeddings to unit interval

Returns:¶

Converted embeddings array

Return type:¶

torch.Tensor or NDArray or T

Raises:¶

• ValueError – If the embeddings array is not 2D

• ValueError – If the embeddings array has a zero dimension

• ValueError – If the embeddings array is not unit interval [0, 1] (when unit_interval=True)