Adaptive Cycle

Returns half the current dominant cycle period — the "best" lookback for downstream oscillators like an adaptive RSI or adaptive Stochastic. Halving accounts for the fact that an oscillator over a half-cycle captures the full peak-to-trough swing without aliasing. Wrapper over HilbertDominantCycle.

Quick reference

Item	Value
Family	Ehlers / Cycle (DSP)
Input type	f64
Output type	f64 (integer-valued, half the Hilbert cycle period)
Output range	[3, 25]
Default parameters	none — AdaptiveCycle::new()
Warmup period	~50 bars (inherited from HilbertDominantCycle)
Interpretation	Use as period argument for adaptive oscillators

Formula

hilbert_period_t = HilbertDominantCycle.update(x_t)
adaptive_t       = round(hilbert_period_t * 0.5).clamp(3, 25)

The clamp [3, 25] matches the typical operating range of period-adaptive oscillators — narrower at the bottom than HilbertDominantCycle's [6, 50] clamp because halving compresses the range. See crates/wickra-core/src/indicators/adaptive_cycle.rs.

Parameters

No parameters. AdaptiveCycle::new() returns a default-constructed estimator; Default is also implemented.

Inputs / Outputs

Indicator<Input = f64, Output = f64>. Output is integer-valued f64 (e.g. 7.0, 12.0). Python: AdaptiveCycle().batch(prices) returns a 1-D np.ndarray. Node: same shape; update(value) returns number | null.

Warmup

Same as HilbertDominantCycle: ~50 bars before usable output, ~100 bars before stable. The wrapper adds no additional state beyond the round/clamp.

Edge cases

• Output is integer-valued. Always a whole number; downstream code can .round() as usize safely.
• Lower clamp at 3. Hilbert reads of [6, 50] become [3, 25] after halving and clamping.
• Constant input. Hilbert reads max period → output reads 25 (the upper clamp).
• Reset. reset() clears the inner Hilbert estimator and the last value.

Examples

Rust

use wickra::{AdaptiveCycle, BatchExt, Indicator};

fn main() {
    let prices: Vec<f64> = (0..300)
        .map(|i| 100.0 + (f64::from(i) * 0.4).sin() * 5.0)
        .collect();
    let mut ac = AdaptiveCycle::new();
    println!("row 200 adaptive period = {:?}", ac.batch(&prices)[200]);
}

Python

import numpy as np
import wickra as ta

prices = 100 + np.sin(np.linspace(0, 120, 300)) * 5
ac = ta.AdaptiveCycle()
print('row 200:', ac.batch(prices)[200])

Node

const wickra = require('wickra');
const ac = new wickra.AdaptiveCycle();
const prices = Array.from({ length: 300 },
  (_, i) => 100 + Math.sin(i * 0.4) * 5);
console.log('row 200:', ac.batch(prices)[200]);

Streaming — feed into adaptive RSI

use wickra::{AdaptiveCycle, Indicator, Rsi};

let mut ac = AdaptiveCycle::new();
let mut rsi_cache: Option<(usize, Rsi)> = None;

let price_stream: Vec<f64> = Vec::new(); // your live price feed
for px in price_stream {
    if let Some(period) = ac.update(px) {
        let p = period as usize;
        // Re-instantiate Rsi when the adaptive period changes
        if rsi_cache.as_ref().map(|(prev, _)| *prev) != Some(p) {
            rsi_cache = Some((p, Rsi::new(p).unwrap()));
        }
        if let Some((_, rsi)) = rsi_cache.as_mut() {
            let _r = rsi.update(px);
        }
    }
}

Interpretation

• Adaptive oscillator period. The output is what period should be for a cycle-aware oscillator that's tracking the current market regime. Feed it into RSI, Stochastic, CCI, or similar.
• Regime indicator. Output stable in the 7-12 band suggests a clean intermediate-cycle market. Output pinned at 3 or 25 suggests no clear cycle.
• Pairs with InverseFisherTransform. A bounded oscillator built on an adaptive-period RSI, with IFT squashing applied, produces a consistently scaled signal across regimes — see Ehlers' cookbook in Cycle Analytics for Traders.

Common pitfalls

• Reusing the same oscillator instance. When the adaptive period changes, an existing oscillator with a fixed period internal buffer can't simply adapt — you must reset or re-instantiate. This is the trickiest part of building adaptive-oscillator systems.
• Treating output as price. It's a period in bars, not a price. Don't plot it on the same chart panel as price.
• Warmup masking. During the ~50-bar warmup, the inner Hilbert output is unstable; downstream adaptive systems should also wait through this period.

References

• John F. Ehlers, Cycle Analytics for Traders, Wiley (2013), ch. 11 — Adaptive Cycle and its use as a period feed.

See also

• HilbertDominantCycle — the underlying period estimator.
• Mama — built on the same phase machinery.
• SineWave — visualises the phase that AdaptiveCycle measures.
• Indicators-Overview — full taxonomy.