Struct RawModel 

pub struct RawModel { /* private fields */ }

Represents a raw model.

Implementations

impl RawModel

pub fn feature_provider(&mut self) -> &mut FeatureProvider

Returns a mutable reference of the feature provider.

pub fn merge(&self) -> Result<MergedModel>

Merges this model and returns MergedModel.

This process integrates the features, so that each edge has three items: a uni-gram cost, a left-connection ID, and a right-connection ID.

Errors

Generated left/right connection ID must be smaller than 2^32.

pub fn unigram_weight_indices(&self) -> &[Option<NonZeroU32>]

Returns the relation between uni-gram feature IDs and weight indices.

pub fn bigram_weight_indices(&self) -> &[HashMap<u32, u32>]

Returns the relation between bi-gram feature IDs and weight indices.

pub fn weights(&self) -> &[f64]

Returns weights.

Trait Implementations

impl<'de, __Context> BorrowDecode<'de, __Context> for RawModel

fn borrow_decode<D: BorrowDecoder<'de, Context = __Context>>( decoder: &mut D, ) -> Result<Self, DecodeError>

Attempt to decode this type with the given BorrowDecode.

impl<Context> Decode<Context> for RawModel

fn decode<D: Decoder>(decoder: &mut D) -> Result<Self, DecodeError>

Attempt to decode this type with the given Decode.

impl Encode for RawModel

fn encode<E: Encoder>(&self, encoder: &mut E) -> Result<(), EncodeError>

Encode a given type.

impl Model for RawModel

fn search_best_path(&self, lattice: &Lattice) -> (Vec<Edge>, f64)

Searches the best path and returns the path and its score.