AdaptiveRsi

Wilder's RSI whose up/down averaging adapts to trendiness via Kaufman's efficiency ratio — fast in clean moves, smooth through chop.

Quick reference

Field	Value
Family	Ehlers / Cycle (DSP)
Input type	f64
Output type	f64
Output range	[0, 100] (50 = neutral)
Default parameters	(period = 14) (Python)
Warmup period	period + 1
Interpretation	> 70 overbought, < 30 oversold; reacts faster in trends.

Formula

ER     = |price_t − price_{t−period}| / Σ |Δprice| over the window   (0..1)
sc     = ( ER·(2/3 − 2/31) + 2/31 )²
avg_gain += sc·(gain − avg_gain),  avg_loss += sc·(loss − avg_loss)
RSI    = 100 · avg_gain / (avg_gain + avg_loss)

A fixed-period RSI must trade responsiveness for smoothness. The adaptive RSI sets its smoothing each bar from Kaufman's efficiency ratio (directional move ÷ total path): in a clean trend (ER ≈ 1) the averages track gains and losses almost immediately; in noise (ER ≈ 0) they barely move, filtering the chop. The 2/3…2/31 bounds are KAMA's standard fast/slow smoothing constants. Source: crates/wickra-core/src/indicators/adaptive_rsi.rs. This is the efficiency-ratio cousin of Ehlers' cycle-adaptive RSI, which instead sets the lookback from the measured dominant cycle.

Parameters

Name	Type	Default	Valid range	Source	Description
period	usize	14 (Python)	>= 1	adaptive_rsi.rs:55	Efficiency-ratio window (and RSI seed length). 0 errors with Error::PeriodZero.

The period getter returns the window; value returns the current output if ready.

Inputs / Outputs

From crates/wickra-core/src/indicators/adaptive_rsi.rs:

use wickra::{Indicator, AdaptiveRsi};
// AdaptiveRsi: Input = f64, Output = f64
const _: fn(&mut AdaptiveRsi, f64) -> Option<f64> = <AdaptiveRsi as Indicator>::update;

An f64 in, an Option<f64> out. The Python binding takes a scalar for update and a 1-D numpy array for batch (NaN warmup); Node takes update(value) and batch(values[]).

Warmup

warmup_period() == period + 1. One bar seeds the previous price; period changes seed the averages; the first value lands on bar period + 1 (first_emission_at_warmup_period pins this).

Edge cases

• Pure uptrend → 100. No losses drive the reading to 100 (pure_uptrend_is_one_hundred pins this).
• Flat market → 50. No gains and no losses give the neutral 50 (flat_market_is_neutral pins this).
• Bounded. The reading stays within [0, 100] (output_in_range pins this).
• Non-finite input. A NaN/∞ input is ignored and the last value returned (ignores_non_finite pins this).
• Reset. r.reset() clears the windows, the averages and the last value (reset_clears_state).

Examples

Rust

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

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut r = AdaptiveRsi::new(5)?;
    let out = r.batch(&(1..=40).map(f64::from).collect::<Vec<_>>());
    println!("{:?}", out.last().unwrap()); // Some(100.0)
    Ok(())
}

Output:

Some(100.0)

Python

import numpy as np
import wickra as ta

r = ta.AdaptiveRsi(14)
x = 100 + np.sin(np.arange(200) * 0.3) * 8
print(r.batch(x)[-1])   # in [0, 100]

Node

const ta = require('wickra');

const r = new ta.AdaptiveRsi(14);
console.log('warmupPeriod:', r.warmupPeriod()); // 15

Streaming

use wickra::{Indicator, AdaptiveRsi};

let mut r = AdaptiveRsi::new(14).unwrap();
let mut last = None;
for i in 0..60 {
    last = r.update(100.0 + (f64::from(i) * 0.3).sin() * 5.0);
}
println!("{last:?}");

Streaming update and batch are equivalent tick-for-tick (batch_equals_streaming pins this).

Interpretation

1. Fewer whipsaws. In chop the smoothing collapses, so the oscillator does not flap across the 30/70 lines the way a fast fixed RSI does.
2. Faster trend reads. When a trend kicks in the efficiency ratio rises and the RSI snaps to overbought/oversold sooner than a long fixed RSI.
3. Same thresholds. Read the 30/50/70 levels exactly as for a standard RSI.

Common pitfalls

• Not a cycle-period RSI. This adapts the smoothing, not the lookback; for the dominant-cycle form pair it with an autocorrelation periodogram.
• ER needs a trend to differ. On purely random data the efficiency ratio sits low and the RSI behaves like a slow fixed one.
• Period meaning. Here period is the efficiency-ratio window, not a Wilder smoothing length.

References

Wilder, J. W. (1978), New Concepts in Technical Trading Systems (RSI); Kaufman, P. J. (1995), Smarter Trading (the efficiency ratio and KAMA smoothing); Ehlers, J. F. (2013), Cycle Analytics for Traders (cycle-adaptive RSI).

See also

• Indicator-Rsi — the fixed-period original.
• Indicator-Kama — Kaufman's adaptive moving average.
• Indicator-AdaptiveCci — the adaptive CCI counterpart.
• Indicators-Overview — the full taxonomy.