Enum PreparedModel 

pub enum PreparedModel {
    Cpu(Box<CpuPlan>),
}

A compiled model prepared for a concrete execution backend.

Variants

Cpu(Box<CpuPlan>)

A model prepared for CPU execution.

Implementations

impl PreparedModel

pub fn required_event_scalars(&self) -> &[String]

Returns the event-scalar columns required by this prepared model.

Requirements are deduplicated while retaining compiled graph order.

pub fn evaluate_batch( &self, params: &ParamValues, batch: &EventBatch, ) -> RuntimeResult<Vec<Complex64>>

Evaluates the model for every event in a batch.

Errors

Returns RuntimeError when parameters or event columns are incompatible, evaluation fails, or a matrix solve is singular.

pub fn evaluate_batch_with_gradient( &self, params: &ParamValues, batch: &EventBatch, ) -> RuntimeResult<Vec<ValueGradient>>

Evaluates the model and its free-parameter gradient for every event in a batch.

Errors

Returns RuntimeError when inputs are incompatible, differentiation or evaluation fails, or the selected backend lacks event-wise gradients.

pub fn prepare( model: &CompiledModel, execution: &Execution, ) -> RuntimeResult<Self>

Prepares a compiled model for the supplied execution context.

Errors

Returns RuntimeError when model lowering, differentiation, backend initialization, or precision selection fails.

pub fn prepare_dataset( &self, execution: &Execution, dataset: &Dataset, ) -> RuntimeResult<PreparedDataset>

Prepares a dataset for repeated evaluation with this model.

Errors

Returns RuntimeError when the dataset cannot be read or cached, its schema is incompatible, or backend preparation fails.

pub fn evaluate_prepared( &self, execution: &Execution, params: &ParamValues, dataset: &PreparedDataset, source: &Dataset, ) -> RuntimeResult<Vec<Complex64>>

Evaluates every event in a prepared dataset while preserving source order.

Backends with a prepared event adapter reuse their retained or streaming prepared blocks. Other backends evaluate the supplied source through the already-selected backend; this operation never substitutes a backend.

Errors

Returns RuntimeError when model and dataset backends differ, source streaming fails, or event evaluation fails.

pub fn evaluate_prepared_many( &self, execution: &Execution, params: &[ParamValues], dataset: &PreparedDataset, source: &Dataset, ) -> RuntimeResult<Vec<Vec<Complex64>>>

Evaluates multiple parameter sets while each prepared event block is active.

Output rows retain parameter-set order and each row retains source-event order.

Errors

Returns RuntimeError when model and dataset backends differ, source streaming fails, or event evaluation fails.

pub fn evaluate_prepared_many_with_reduction( &self, execution: &Execution, params: &[ParamValues], dataset: &PreparedDataset, source: &Dataset, reduction: ReductionPlan, ) -> RuntimeResult<(Vec<Vec<Complex64>>, Vec<f64>)>

Evaluates and reduces multiple parameter sets while each prepared block is active.

Errors

Returns RuntimeError when model and dataset backends differ, source streaming fails, event evaluation fails, or the reduction rejects a value.

pub fn reduce( &self, execution: &Execution, params: &ParamValues, dataset: &PreparedDataset, reduction: ReductionPlan, ) -> RuntimeResult<f64>

Executes a weighted scalar reduction over a prepared dataset.

Errors

Returns RuntimeError when model and dataset backends differ, inputs are incompatible, evaluation fails, or the reduction domain is invalid.

pub fn reduce_with_gradient( &self, execution: &Execution, params: &ParamValues, dataset: &PreparedDataset, reduction: ReductionPlan, ) -> RuntimeResult<ReductionEvaluation>

Executes a weighted reduction and computes its free-parameter gradient.

Errors

Returns RuntimeError when model and dataset backends differ, differentiation or evaluation fails, or the reduction domain is invalid.

Trait Implementations

impl Clone for PreparedModel

fn clone(&self) -> PreparedModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PreparedModel

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

Formats the value using the given formatter.