SmoothedHeikinAshi

The Heikin-Ashi transform applied to EMA-smoothed OHLC — long, clean runs of same-colour candles for an even clearer trend read.

Quick reference

Field	Value
Family	Ichimoku & Charts
Input type	Candle (open / high / low / close)
Output type	SmoothedHeikinAshiOutput { open, high, low, close }
Output range	price units; low <= open,close <= high
Default parameters	(period = 10) (Python)
Warmup period	period
Interpretation	close > open bullish candle; long runs = strong trend.

Formula

eo, eh, el, ec = EMA(open|high|low|close, period)
ha_close = (eo + eh + el + ec) / 4
ha_open  = (prev_ha_open + prev_ha_close) / 2     (seeded with (eo + ec)/2)
ha_high  = max(eh, ha_open, ha_close)
ha_low   = min(el, ha_open, ha_close)

Standard Heikin-Ashi averages the OHLC; this variant first smooths each input series with an EMA, removing additional noise so trends print as uninterrupted colour runs with crisp flips at turns — at the cost of period-proportional lag. Source: crates/wickra-core/src/indicators/smoothed_heikin_ashi.rs.

Parameters

Name	Type	Default	Valid range	Source	Description
period	usize	10 (Python)	>= 1	smoothed_heikin_ashi.rs:69	EMA smoothing length applied to each OHLC series. 0 errors with Error::PeriodZero.

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

Inputs / Outputs

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

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

A Candle in, an Option<SmoothedHeikinAshiOutput> out. The Python binding returns an (open, high, low, close) tuple from update and an (n, 4) array from batch(open, high, low, close); Node returns { open, high, low, close } and a flat Float64Array of length n*4; WASM mirrors the object with camelCase keys.

Warmup

warmup_period() == period. The four EMAs seed before the first candle (first_emission_at_warmup_period pins this).

Edge cases

• Bracketing. high >= open,close and low <= open,close (high_brackets_open_close pins this).
• Uptrend → bullish candle. A clean rise prints close > open (uptrend_close_above_open pins this).
• Finiteness. Candle::new rejects non-finite fields, so no in-method guard is needed.
• Reset. s.reset() clears the four EMAs, the prior candle and the last value (reset_clears_state).

Examples

Rust

use wickra::{BatchExt, Candle, Indicator, SmoothedHeikinAshi};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut s = SmoothedHeikinAshi::new(3)?;
    let candles: Vec<Candle> = (0..30)
        .map(|i| { let b = 100.0 + 2.0 * f64::from(i); Candle::new(b, b + 1.0, b - 1.0, b + 0.5, 1_000.0, 0).unwrap() })
        .collect();
    let out = s.batch(&candles).last().unwrap().unwrap();
    println!("bullish candle: {}", out.close > out.open);
    Ok(())
}

Output:

bullish candle: true

Python

import numpy as np
import wickra as ta

s = ta.SmoothedHeikinAshi(10)
o, h, l, c = (np.arange(40.0)+100, np.arange(40.0)+101, np.arange(40.0)+99, np.arange(40.0)+100.5)
ha = s.batch(o, h, l, c)  # (n, 4)
print(ha[-1])

Node

const ta = require('wickra');

const s = new ta.SmoothedHeikinAshi(10);
console.log('warmupPeriod:', s.warmupPeriod()); // 10

Streaming

use wickra::{Candle, Indicator, SmoothedHeikinAshi};

let mut s = SmoothedHeikinAshi::new(10).unwrap();
let feed: Vec<Candle> = Vec::new(); // your live stream
for candle in feed {
    if let Some(ha) = s.update(candle) {
        // ha.close > ha.open -> bullish; colour flip -> trend change
    }
}

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

Interpretation

1. Trend persistence. A run of bullish (close > open) candles with small lower wicks signals a strong, low-noise uptrend (mirror for downtrends).
2. Colour flip. The first opposite-colour candle after a long run is the earliest visual reversal cue.
3. Lag awareness. Bigger period = smoother but later flips; tune to your horizon.

Common pitfalls

• Not real prices. Smoothed HA OHLC are synthetic; never use the levels for exact entries/stops — use them for direction.
• Lag. The double smoothing delays reversals more than plain HA.
• Repaint-free here. This is computed bar-closed; intraday updates will move the forming candle (as with any HA).

References

Heikin-Ashi originates in Japanese candlestick practice; the smoothed variant is a common platform extension (e.g. Dan Valcu's work on Heikin-Ashi).

See also

• Indicator-HeikinAshi — the unsmoothed transform.
• Indicator-HeikinAshiOscillator — the HA body as an oscillator.
• Indicator-Ema — the smoothing core.
• Indicators-Overview — the full taxonomy.