Struct FSRS

pub struct FSRS<B: Backend = NdArray> { /* private fields */ }

This is the main structure provided by this crate. It can be used for both parameter training, and for reviews.

Implementations

impl<B: Backend> FSRS<B>

pub fn memory_state( &self, item: FSRSItem, starting_state: Option<MemoryState>, ) -> Result<MemoryState>

Calculate the current memory state for a given card’s history of reviews. In the case of truncated reviews, starting_state can be set to the value of FSRS::memory_state_from_sm2 for the first review (which should not be included in FSRSItem). If not provided, the card starts as new. Parameters must have been provided when calling FSRS::new().

pub fn memory_state_from_sm2( &self, ease_factor: f32, interval: f32, sm2_retention: f32, ) -> Result<MemoryState>

If a card has incomplete learning history, memory state can be approximated from current sm2 values. Parameters must have been provided when calling FSRS::new().

pub fn next_interval( &self, stability: Option<f32>, desired_retention: f32, rating: u32, ) -> f32

Calculate the next interval for the current memory state, for rescheduling. Stability should be provided except when the card is new. Rating is ignored except when card is new. Parameters must have been provided when calling FSRS::new().

pub fn next_states( &self, current_memory_state: Option<MemoryState>, desired_retention: f32, days_elapsed: u32, ) -> Result<NextStates>

The intervals and memory states for each answer button. Parameters must have been provided when calling FSRS::new().

pub fn evaluate<F>( &self, items: Vec<FSRSItem>, progress: F, ) -> Result<ModelEvaluation>

where F: FnMut(ItemProgress) -> bool,

Determine how well the model and parameters predict performance. Parameters must have been provided when calling FSRS::new().

pub fn current_retrievability( &self, state: MemoryState, days_elapsed: u32, ) -> f32

How well the user is likely to remember the item after days_elapsed since the previous review.

pub fn universal_metrics<F>( &self, items: Vec<FSRSItem>, parameters: &[f32], progress: F, ) -> Result<(f32, f32)>

where F: FnMut(ItemProgress) -> bool,

Returns the universal metrics for the existing and provided parameters. If the first value is smaller than the second value, the existing parameters are better than the provided ones.

impl FSRS<NdArray>

pub fn new(parameters: Option<&[f32]>) -> Result<Self>

• Parameters must be provided before running commands that need them.
• Parameters may be an empty slice to use the default values instead.

impl<B: Backend> FSRS<B>

pub fn new_with_backend<B2: Backend>( parameters: Option<&[f32]>, device: B2::Device, ) -> Result<FSRS<B2>>

impl<B: Backend> FSRS<B>

pub fn optimal_retention<F>( &self, config: &SimulatorConfig, parameters: &[f32], progress: F, ) -> Result<f32>

where F: FnMut(ItemProgress) -> bool + Send,

For the given simulator parameters and parameters, determine the suggested desired_retention value.

impl<B: Backend> FSRS<B>

pub fn compute_parameters( &self, train_set: Vec<FSRSItem>, progress: Option<Arc<Mutex<CombinedProgressState>>>, enable_short_term: bool, ) -> Result<Vec<f32>>

Calculate appropriate parameters for the provided review history.

pub fn benchmark( &self, train_set: Vec<FSRSItem>, enable_short_term: bool, ) -> Vec<f32>

Trait Implementations

impl<B: Clone + Backend> Clone for FSRS<B>

where B::Device: Clone,

fn clone(&self) -> FSRS<B>

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<B: Debug + Backend> Debug for FSRS<B>

where B::Device: Debug,

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

Formats the value using the given formatter.