nautilus-indicators 0.63.0

// -------------------------------------------------------------------------------------------------
//  Copyright (C) 2015-2026 Nautech Systems Pty Ltd. All rights reserved.
//  https://nautechsystems.io
//
//  Licensed under the GNU Lesser General Public License Version 3.0 (the "License");
//  You may not use this file except in compliance with the License.
//  You may obtain a copy of the License at https://www.gnu.org/licenses/lgpl-3.0.en.html
//
//  Unless required by applicable law or agreed to in writing, software
//  distributed under the License is distributed on an "AS IS" BASIS,
//  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
//  See the License for the specific language governing permissions and
//  limitations under the License.
// -------------------------------------------------------------------------------------------------

use std::fmt::Display;

use nautilus_core::correctness::{FAILED, check_predicate_true};
use nautilus_model::{
    data::{Bar, QuoteTick, TradeTick},
    enums::PriceType,
};

use crate::{
    indicator::{Indicator, MovingAverage},
    ratio::efficiency_ratio::EfficiencyRatio,
};

/// An indicator which calculates an adaptive moving average (AMA) across a
/// rolling window. Developed by Perry Kaufman, the AMA is a moving average
/// designed to account for market noise and volatility. The AMA will closely
/// follow prices when the price swings are relatively small and the noise is
/// low. The AMA will increase lag when the price swings increase.
#[repr(C)]
#[derive(Debug)]
#[cfg_attr(
    feature = "python",
    pyo3::pyclass(module = "nautilus_trader.indicators")
)]
#[cfg_attr(
    feature = "python",
    pyo3_stub_gen::derive::gen_stub_pyclass(module = "nautilus_trader.indicators")
)]
pub struct AdaptiveMovingAverage {
    /// The period for the internal `EfficiencyRatio` indicator (>= 2).
    pub period_efficiency_ratio: usize,
    /// The period for the fast smoothing constant (> 0).
    pub period_fast: usize,
    /// The period for the slow smoothing constant (> `period_fast`).
    pub period_slow: usize,
    /// The price type used for calculations.
    pub price_type: PriceType,
    /// The last indicator value.
    pub value: f64,
    /// The input count for the indicator.
    pub count: usize,
    pub initialized: bool,
    has_inputs: bool,
    efficiency_ratio: EfficiencyRatio,
    prior_value: Option<f64>,
    alpha_fast: f64,
    alpha_slow: f64,
}

impl Display for AdaptiveMovingAverage {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        write!(
            f,
            "{}({},{},{})",
            self.name(),
            self.period_efficiency_ratio,
            self.period_fast,
            self.period_slow
        )
    }
}

impl Indicator for AdaptiveMovingAverage {
    fn name(&self) -> String {
        stringify!(AdaptiveMovingAverage).to_string()
    }

    fn has_inputs(&self) -> bool {
        self.has_inputs
    }

    fn initialized(&self) -> bool {
        self.initialized
    }

    fn handle_quote(&mut self, quote: &QuoteTick) -> anyhow::Result<()> {
        self.update_raw(quote.extract_price(self.price_type)?.into());
        Ok(())
    }

    fn handle_trade(&mut self, trade: &TradeTick) {
        self.update_raw((&trade.price).into());
    }

    fn handle_bar(&mut self, bar: &Bar) {
        self.update_raw((&bar.close).into());
    }

    fn reset(&mut self) {
        self.value = 0.0;
        self.prior_value = None;
        self.count = 0;
        self.has_inputs = false;
        self.initialized = false;
        self.efficiency_ratio.reset();
    }
}

impl AdaptiveMovingAverage {
    /// Creates a new [`AdaptiveMovingAverage`] instance.
    ///
    /// # Panics
    ///
    /// This function panics if:
    /// - `period_efficiency_ratio` is less than 2 or its rolling-window storage
    ///   cannot be reserved.
    /// - `period_fast` == 0.
    /// - `period_slow` == 0.
    /// - `period_slow` == `usize::MAX`.
    /// - `period_slow` ≤ `period_fast`.
    #[must_use]
    pub fn new(
        period_efficiency_ratio: usize,
        period_fast: usize,
        period_slow: usize,
        price_type: Option<PriceType>,
    ) -> Self {
        Self::new_checked(
            period_efficiency_ratio,
            period_fast,
            period_slow,
            price_type,
        )
        .expect(FAILED)
    }

    pub(crate) fn new_checked(
        period_efficiency_ratio: usize,
        period_fast: usize,
        period_slow: usize,
        price_type: Option<PriceType>,
    ) -> anyhow::Result<Self> {
        check_predicate_true(period_fast > 0, "`period_fast` must be positive")?;
        check_predicate_true(period_slow > 0, "`period_slow` must be positive")?;
        check_predicate_true(
            period_slow < usize::MAX,
            "`period_slow` must be less than `usize::MAX`",
        )?;
        check_predicate_true(
            period_slow > period_fast,
            "`period_slow` must be greater than `period_fast`",
        )?;

        let efficiency_ratio = EfficiencyRatio::new_checked(period_efficiency_ratio, price_type)?;

        Ok(Self {
            period_efficiency_ratio,
            period_fast,
            period_slow,
            price_type: price_type.unwrap_or(PriceType::Last),
            value: 0.0,
            count: 0,
            alpha_fast: 2.0 / (period_fast + 1) as f64,
            alpha_slow: 2.0 / (period_slow + 1) as f64,
            prior_value: None,
            has_inputs: false,
            initialized: false,
            efficiency_ratio,
        })
    }

    #[must_use]
    pub fn alpha_diff(&self) -> f64 {
        self.alpha_fast - self.alpha_slow
    }

    #[must_use]
    pub const fn alpha_fast(&self) -> f64 {
        self.alpha_fast
    }

    #[must_use]
    pub const fn alpha_slow(&self) -> f64 {
        self.alpha_slow
    }

    pub fn reset(&mut self) {
        Indicator::reset(self);
    }
}

impl MovingAverage for AdaptiveMovingAverage {
    fn value(&self) -> f64 {
        self.value
    }

    fn count(&self) -> usize {
        self.count
    }

    fn update_raw(&mut self, value: f64) {
        self.count += 1;

        if !self.has_inputs {
            self.prior_value = Some(value);
            self.efficiency_ratio.update_raw(value);
            self.value = value;
            self.has_inputs = true;
            return;
        }

        self.efficiency_ratio.update_raw(value);
        self.prior_value = Some(self.value);

        // Calculate the smoothing constant
        let smoothing_constant = self
            .efficiency_ratio
            .value
            .mul_add(self.alpha_diff(), self.alpha_slow)
            .powi(2);

        // Calculate the AMA
        // TODO: Remove unwraps
        self.value = smoothing_constant
            .mul_add(value - self.prior_value.unwrap(), self.prior_value.unwrap());

        if self.efficiency_ratio.initialized() {
            self.initialized = true;
        }
    }
}

#[cfg(test)]
mod tests {
    use nautilus_model::data::{Bar, QuoteTick, TradeTick};
    use rstest::rstest;

    use crate::{
        average::ama::AdaptiveMovingAverage,
        indicator::{Indicator, MovingAverage},
        stubs::*,
        testing::assert_approx_equal,
    };

    #[rstest]
    fn test_ama_initialized(indicator_ama_10: AdaptiveMovingAverage) {
        let display_str = format!("{indicator_ama_10}");
        assert_eq!(display_str, "AdaptiveMovingAverage(10,2,30)");
        assert_eq!(indicator_ama_10.name(), "AdaptiveMovingAverage");
        assert!(!indicator_ama_10.has_inputs());
        assert!(!indicator_ama_10.initialized());
    }

    #[rstest]
    fn test_value_with_one_input(mut indicator_ama_10: AdaptiveMovingAverage) {
        indicator_ama_10.update_raw(1.0);
        assert_eq!(indicator_ama_10.value, 1.0);
    }

    #[rstest]
    fn test_value_with_two_inputs(mut indicator_ama_10: AdaptiveMovingAverage) {
        indicator_ama_10.update_raw(1.0);
        indicator_ama_10.update_raw(2.0);
        assert_approx_equal(indicator_ama_10.value, 1.44444444444);
    }

    #[rstest]
    fn test_value_with_three_inputs(mut indicator_ama_10: AdaptiveMovingAverage) {
        indicator_ama_10.update_raw(1.0);
        indicator_ama_10.update_raw(2.0);
        indicator_ama_10.update_raw(3.0);
        assert_approx_equal(indicator_ama_10.value, 2.13580246914);
    }

    #[rstest]
    #[case::inherent(AdaptiveMovingAverage::reset)]
    #[case::indicator(<AdaptiveMovingAverage as Indicator>::reset)]
    fn test_reset(
        #[case] reset: fn(&mut AdaptiveMovingAverage),
        mut indicator_ama_10: AdaptiveMovingAverage,
    ) {
        for value in 1..=10 {
            indicator_ama_10.update_raw(f64::from(value));
        }
        assert!(indicator_ama_10.initialized);

        reset(&mut indicator_ama_10);

        assert!(!indicator_ama_10.initialized);
        assert!(!indicator_ama_10.has_inputs);
        assert_eq!(indicator_ama_10.value, 0.0);
        assert_eq!(indicator_ama_10.prior_value, None);
        assert_eq!(indicator_ama_10.count, 0);
        assert!(!indicator_ama_10.efficiency_ratio.has_inputs());
        assert!(!indicator_ama_10.efficiency_ratio.initialized());
        assert_eq!(indicator_ama_10.efficiency_ratio.value, 0.0);
    }

    #[rstest]
    fn test_initialized_after_correct_number_of_input(indicator_ama_10: AdaptiveMovingAverage) {
        let mut ama = indicator_ama_10;
        for _ in 0..9 {
            ama.update_raw(1.0);
        }
        assert!(!ama.initialized);
        ama.update_raw(1.0);
        assert!(ama.initialized);
    }

    #[rstest]
    fn test_count_increments(mut indicator_ama_10: AdaptiveMovingAverage) {
        assert_eq!(indicator_ama_10.count(), 0);
        indicator_ama_10.update_raw(1.0);
        assert_eq!(indicator_ama_10.count(), 1);
        indicator_ama_10.update_raw(2.0);
        indicator_ama_10.update_raw(3.0);
        assert_eq!(indicator_ama_10.count(), 3);
    }

    #[rstest]
    fn test_handle_quote_tick(mut indicator_ama_10: AdaptiveMovingAverage, stub_quote: QuoteTick) {
        indicator_ama_10.handle_quote(&stub_quote).unwrap();
        assert!(indicator_ama_10.has_inputs);
        assert!(!indicator_ama_10.initialized);
        assert_eq!(indicator_ama_10.value, 1501.0);
        assert_eq!(indicator_ama_10.count(), 1);
    }

    #[rstest]
    fn test_handle_trade_tick_update(
        mut indicator_ama_10: AdaptiveMovingAverage,
        stub_trade: TradeTick,
    ) {
        indicator_ama_10.handle_trade(&stub_trade);
        assert!(indicator_ama_10.has_inputs);
        assert!(!indicator_ama_10.initialized);
        assert_eq!(indicator_ama_10.value, 1500.0);
        assert_eq!(indicator_ama_10.count(), 1);
    }

    #[rstest]
    fn handle_handle_bar(
        mut indicator_ama_10: AdaptiveMovingAverage,
        bar_ethusdt_binance_minute_bid: Bar,
    ) {
        indicator_ama_10.handle_bar(&bar_ethusdt_binance_minute_bid);
        assert!(indicator_ama_10.has_inputs);
        assert!(!indicator_ama_10.initialized);
        assert_eq!(indicator_ama_10.value, 1522.0);
        assert_eq!(indicator_ama_10.count(), 1);
    }

    #[rstest]
    fn new_panics_when_slow_not_greater_than_fast() {
        let result = std::panic::catch_unwind(|| {
            let _ = AdaptiveMovingAverage::new(10, 20, 20, None);
        });
        assert!(result.is_err());
    }

    #[rstest]
    #[case(0)]
    #[case(1)]
    #[should_panic(expected = "`period` must be at least 2")]
    fn new_panics_when_er_period_is_below_two(#[case] period: usize) {
        let _ = AdaptiveMovingAverage::new(period, 2, 30, None);
    }

    #[rstest]
    fn new_panics_when_fast_is_zero() {
        let result = std::panic::catch_unwind(|| {
            let _ = AdaptiveMovingAverage::new(10, 0, 30, None);
        });
        assert!(result.is_err());
    }

    #[rstest]
    fn new_panics_when_slow_is_zero() {
        let result = std::panic::catch_unwind(|| {
            let _ = AdaptiveMovingAverage::new(10, 2, 0, None);
        });
        assert!(result.is_err());
    }

    #[rstest]
    fn new_panics_when_slow_less_than_fast() {
        let result = std::panic::catch_unwind(|| {
            let _ = AdaptiveMovingAverage::new(10, 20, 5, None);
        });
        assert!(result.is_err());
    }

    #[rstest]
    fn new_checked_rejects_slow_period_max() {
        let error =
            AdaptiveMovingAverage::new_checked(10, usize::MAX - 1, usize::MAX, None).unwrap_err();

        assert_eq!(
            error.to_string(),
            "`period_slow` must be less than `usize::MAX`",
        );
    }
}