Enum WeightType 

pub enum WeightType {
    F64,
    F32,
}

Edge-weight storage width of a graph collection, mirroring NodeIdWidth (#106): the width is schema-declared through GraphWriteOptions::weight_type, resolving the RFC #81 open question (“Float64 only, or allow Float32?”) with “both, schema-declared”.

Variants

F64

Float64 weights (default): the same width and exactness as the value column of the legacy sparse-matrix artifacts, so an f64 CSR (laplacian, adjacency) round-trips bit-identically through the graph API just like through save_sparse.

F32

Float32 weights (opt-in): half the storage bytes for memory-bound consumers. Weights that cannot be represented exactly at this width are rejected on save instead of being silently narrowed; finite values beyond the f32 range surface StorageError::Overflow (#51). Infinities narrow bit-exactly; NaN is preserved as NaN (classification semantics only — the payload is not guaranteed bit-exact, unlike the WeightType::F64 width).

Implementations

impl WeightType

pub const fn as_str(&self) -> &'static str

pub fn parse(s: &str) -> StorageResult<Self>

Trait Implementations

impl Clone for WeightType

fn clone(&self) -> WeightType

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for WeightType

impl Debug for WeightType

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

Formats the value using the given formatter.

impl Eq for WeightType

impl PartialEq for WeightType

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for WeightType