Crate fitts 

Fitts’ Law Spaced Repetition Library

A hybrid cognitive model combining:

• SM-2: Classic spaced repetition algorithm for interval scheduling
• Fitts’ Law: Response time prediction with adaptive calibration

Key Innovation: Dual Signal Capture

Unlike traditional flashcard systems that only capture rating, this library captures two complementary signals:

Signal	Type	Source	Purpose
Rating	Subjective	User input	“How well did I recall?” → SM-2
Response Time	Objective	Measured	“How fast did I recall?” → Fitts calibration

This enables personalized predictions that adapt to each user’s patterns.

Quick Start

use fitts::{FittsScheduler, CardState, Rating, ReviewInput};

let mut scheduler = FittsScheduler::new();
let card = CardState::default();

// Option 1: Rating only (traditional)
let result = scheduler.review(card.clone(), Rating::Good);

// Option 2: Rating + Response Time (enables calibration)
let input = ReviewInput::new(Rating::Good, 2500); // 2.5 seconds
let result = scheduler.review(card, input);

println!("Next review in {} days", result.card.interval_days);
println!("Predicted RT: {:.1}s", result.predicted_rt);
if let Some(actual) = result.actual_rt {
    println!("Actual RT: {:.1}s", actual);
}

Adaptive Calibration

Based on ACT-R (Anderson, 1993) and Pavlik & Anderson (2008):

error = RT_actual - RT_predicted
a_new = a + α × error
b_new = b + α × error × log₂(ID + 1)

The model learns each user’s response patterns over time.

Algorithm Details

SM-2 Algorithm (Wozniak, 1987)

Intervals: 1 day → 6 days → previous × EF

Rating → Quality mapping:

• Again = quality 1 (failure, reset)
• Hard = quality 3 (success, maintain)
• Good = quality 4 (success, progress)
• Easy = quality 5 (success, accelerate)

Fitts’ Law Adaptation

Original (1954): RT = a + b × log₂(2D/W)

Memory adaptation:

• Distance = memory decay (time × difficulty)
• Width = accessibility (ease × stability)

Formula: RT = a + b × log₂(distance/accessibility + 1)

Structs

CalibrationResult

Result of a calibration step.

CardState

Card state for SM-2 scheduling.

Deck

A deck of flashcards.

DeckStats

Deck statistics.

FittsModel

Fitts’ Law memory model with adaptive calibration.

FittsParameters

Fitts model parameters.

FittsScheduler

SM-2 scheduler with adaptive Fitts’ Law predictions.

ReviewInput

Input for a card review.

ReviewResult

Result of a card review.

ScheduledCard

Card scheduled for review with predictions.

Scheduler

Main scheduler for managing decks and reviews.

SchedulerConfig

Scheduler configuration.

Enums

DifficultyLevel

Predicted difficulty level based on Fitts’ Law response time.

FittsError

Errors in Fitts model operations.

Rating

User rating for a review (subjective input from user).

SchedulerError

Scheduler errors.