Struct QuantumDDPG 

pub struct QuantumDDPG { /* private fields */ }

Quantum Deep Deterministic Policy Gradient (QDDPG)

Implementations

impl QuantumDDPG

pub fn new( state_dim: usize, action_dim: usize, action_bounds: Vec<(f64, f64)>, num_qubits: usize, buffer_capacity: usize, ) -> Result<Self>

Create new QDDPG agent

Examples found in repository

examples/continuous_rl.rs (lines 77-83)

fn train_qddpg_pendulum() -> Result<()> {
    let state_dim = 3;
    let action_dim = 1;
    let action_bounds = vec![(-2.0, 2.0)];
    let num_qubits = 4;
    let buffer_capacity = 10000;

    // Create QDDPG agent
    let mut agent = QuantumDDPG::new(
        state_dim,
        action_dim,
        action_bounds,
        num_qubits,
        buffer_capacity,
    )?;

    // Create environment
    let mut env = PendulumEnvironment::new();

    // Create optimizers
    let mut actor_optimizer = Adam::new(0.001);
    let mut critic_optimizer = Adam::new(0.001);

    // Train for a few episodes (reduced for demo)
    let episodes = 50;
    println!("   Training QDDPG for {episodes} episodes...");

    let rewards = agent.train(
        &mut env,
        episodes,
        &mut actor_optimizer,
        &mut critic_optimizer,
    )?;

    // Print training statistics
    let avg_initial = rewards[..10].iter().sum::<f64>() / 10.0;
    let avg_final = rewards[rewards.len() - 10..].iter().sum::<f64>() / 10.0;

    println!("\n   Training Statistics:");
    println!("   - Average initial reward: {avg_initial:.2}");
    println!("   - Average final reward: {avg_final:.2}");
    println!("   - Improvement: {:.2}", avg_final - avg_initial);

    // Test trained agent
    println!("\n   Testing trained agent...");
    test_trained_agent(&agent, &mut env)?;

    Ok(())
}

pub fn get_action( &self, state: &Array1<f64>, training: bool, ) -> Result<Array1<f64>>

Get action for state

Examples found in repository

examples/continuous_rl.rs (line 131)

fn test_trained_agent(agent: &QuantumDDPG, env: &mut dyn ContinuousEnvironment) -> Result<()> {
    let test_episodes = 5;
    let mut test_rewards = Vec::new();

    for episode in 0..test_episodes {
        let mut state = env.reset();
        let mut episode_reward = 0.0;
        let mut done = false;
        let mut steps = 0;

        while !done && steps < 200 {
            let action = agent.get_action(&state, false)?; // No exploration
            let (next_state, reward, is_done) = env.step(action.clone())?;

            state = next_state;
            episode_reward += reward;
            done = is_done;
            steps += 1;
        }

        test_rewards.push(episode_reward);
        println!(
            "   Test episode {}: Reward = {:.2}, Steps = {}",
            episode + 1,
            episode_reward,
            steps
        );
    }

    let avg_test = test_rewards.iter().sum::<f64>() / f64::from(test_episodes);
    println!("   Average test reward: {avg_test:.2}");

    Ok(())
}

pub fn store_experience(&mut self, exp: Experience)

Store experience in replay buffer

pub fn update( &mut self, actor_optimizer: &mut dyn Optimizer, critic_optimizer: &mut dyn Optimizer, ) -> Result<()>

Update networks

pub fn train( &mut self, env: &mut dyn ContinuousEnvironment, episodes: usize, actor_optimizer: &mut dyn Optimizer, critic_optimizer: &mut dyn Optimizer, ) -> Result<Vec<f64>>

Train on environment

Examples found in repository

examples/continuous_rl.rs (lines 96-101)

fn train_qddpg_pendulum() -> Result<()> {
    let state_dim = 3;
    let action_dim = 1;
    let action_bounds = vec![(-2.0, 2.0)];
    let num_qubits = 4;
    let buffer_capacity = 10000;

    // Create QDDPG agent
    let mut agent = QuantumDDPG::new(
        state_dim,
        action_dim,
        action_bounds,
        num_qubits,
        buffer_capacity,
    )?;

    // Create environment
    let mut env = PendulumEnvironment::new();

    // Create optimizers
    let mut actor_optimizer = Adam::new(0.001);
    let mut critic_optimizer = Adam::new(0.001);

    // Train for a few episodes (reduced for demo)
    let episodes = 50;
    println!("   Training QDDPG for {episodes} episodes...");

    let rewards = agent.train(
        &mut env,
        episodes,
        &mut actor_optimizer,
        &mut critic_optimizer,
    )?;

    // Print training statistics
    let avg_initial = rewards[..10].iter().sum::<f64>() / 10.0;
    let avg_final = rewards[rewards.len() - 10..].iter().sum::<f64>() / 10.0;

    println!("\n   Training Statistics:");
    println!("   - Average initial reward: {avg_initial:.2}");
    println!("   - Average final reward: {avg_final:.2}");
    println!("   - Improvement: {:.2}", avg_final - avg_initial);

    // Test trained agent
    println!("\n   Testing trained agent...");
    test_trained_agent(&agent, &mut env)?;

    Ok(())
}