Struct NdTensorInput 

pub struct NdTensorInput {
    pub tensor_id: String,
    pub representation_id: RepresentationId,
    pub container: String,
    pub dtype: NdTensorDType,
    pub shape: Vec<usize>,
    pub observation_ids: Vec<ObservationId>,
    pub sample_ids: Option<Vec<SampleId>>,
    pub data: Vec<u8>,
    pub row_presence: Option<Vec<bool>>,
}

A canonical (contiguous row-major) N-D tensor handed to the store. The C ABI layer copies any strided borrowed input into this shape before calling in.

Fields

tensor_id: String

representation_id: RepresentationId

container: String

dtype: NdTensorDType

shape: Vec<usize>

Full shape; shape[0] is the sample/observation axis.

observation_ids: Vec<ObservationId>

One id per axis-0 row; len == shape[0].

sample_ids: Option<Vec<SampleId>>

Optional advisory sample ids (validated for length only; the authoritative sample mapping comes from the coordinator relations at export time).

data: Vec<u8>

Contiguous row-major bytes; len == product(shape) * dtype.element_size().

Multibyte elements (F64/F32/I32) MUST be encoded little-endian. This layer copies the bytes verbatim and never reinterprets them as native-endian values, so the byte order is part of the input contract, not something the layer can fix up. The content fingerprint hashes these bytes as-is; pinning little-endian here is what makes the fingerprint platform-independent (a big-endian host must canonicalize to LE before constructing the input).

row_presence: Option<Vec<bool>>

Optional per-axis-0-row presence mask; len == shape[0].

Trait Implementations

impl Clone for NdTensorInput

fn clone(&self) -> NdTensorInput

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for NdTensorInput

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl PartialEq for NdTensorInput

fn eq(&self, other: &NdTensorInput) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for NdTensorInput