AbandonedBaby

Strong three-bar reversal where a doji is "abandoned" by price gaps on both sides, isolating it from the candle before and after. One of the rarer and more reliable classical reversals.

Quick reference

Field	Value
Family	Candlestick Patterns
Input type	Candle
Output type	f64 — +1.0 bullish, -1.0 bearish, 0.0 otherwise
Output range	{-1.0, 0.0, +1.0}
Default parameters	tolerance = 0.001 (doji flatness); bindings use the default
Warmup period	3 (first two bars always 0.0)
Interpretation	High-conviction reversal; gap-isolated doji marks the turn

Formula

tol = tolerance · max(|bar2.open|, |bar2.close|)
bar2 doji:  |bar2.close − bar2.open| <= tol

bullish (+1.0): bar1 red, bar2 gaps fully below bar1 (bar2.high < bar1.low),
                bar3 green and gaps fully above bar2 (bar3.low > bar2.high)
bearish (-1.0): bar1 green, bar2 gaps fully above bar1 (bar2.low > bar1.high),
                bar3 red and gaps fully below bar2 (bar3.high < bar2.low)

The gaps are full (shadow-to-shadow), not just body gaps — that is what makes the doji truly abandoned. See crates/wickra-core/src/indicators/abandoned_baby.rs.

Parameters

Name	Type	Default	Valid range	Source
tolerance	f64	0.001	[0.0, 1.0)	AbandonedBaby::with_tolerance (abandoned_baby.rs)

with_tolerance outside [0, 1) errors (rejects_invalid_tolerance, accepts_valid_tolerance). Python/Node construct with the default.

Signed ±1 encoding

Emits the uniform candlestick sign convention — +1.0 bullish, −1.0 bearish, 0.0 no pattern — a single feature-matrix dimension.

Inputs / Outputs

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

• Always emits a value. Never None; warmup and no-match bars return 0.0.
• Node. update(open, high, low, close) → number; batch(open, high, low, close) → Array<number>.
• Python. update(candle) → float; batch(open, high, low, close) → 1-D array.array('d') (0.0 on warmup / no-match).

Warmup

warmup_period() == 3. The first two bars return 0.0 (first_two_bars_return_zero, accessors_and_metadata).

Edge cases

• Middle must be a doji. A non-doji middle candle yields 0.0 (middle_not_doji_yields_zero).
• Both gaps required. Missing either full gap yields 0.0 (no_gap_yields_zero).
• Reset. reset() clears the two-bar cache (reset_clears_state).

Examples

Rust

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

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut t = AbandonedBaby::new();
    println!("{:?}", t.update(Candle::new(20.0, 20.1, 14.9, 15.0, 1.0, 0)?)); // red
    println!("{:?}", t.update(Candle::new(13.0, 13.1, 12.9, 13.0, 1.0, 1)?)); // doji, gaps below
    println!("{:?}", t.update(Candle::new(16.0, 18.1, 15.9, 18.0, 1.0, 2)?)); // green, gaps above
    Ok(())
}

Output:

Some(0.0)
Some(0.0)
Some(1.0)

The doji's high 13.1 is below bar1's low 14.9, and bar3's low 15.9 is above the doji's high 13.1 — fully abandoned. This matches bullish_abandoned_baby_is_plus_one.

Python

import numpy as np
import wickra as ta

o = np.array([20.0, 13.0, 16.0])
h = np.array([20.1, 13.1, 18.1])
l = np.array([14.9, 12.9, 15.9])
c = np.array([15.0, 13.0, 18.0])

print(ta.AbandonedBaby().batch(o, h, l, c))  # [0. 0. 1.]

Node

const ta = require('wickra');
const t = new ta.AbandonedBaby();
t.update(20, 20.1, 14.9, 15);
t.update(13, 13.1, 12.9, 13);
console.log(t.update(16, 18.1, 15.9, 18)); // 1

Streaming

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

let mut t = AbandonedBaby::new();
let candle_stream: Vec<wickra::Candle> = Vec::new(); // your live OHLCV candle feed
for bar in candle_stream {
    match t.update(bar) {
        Some(1.0)  => { /* bullish island reversal */ }
        Some(-1.0) => { /* bearish island reversal */ }
        _ => {}
    }
}

Interpretation

1. Island reversal. The doji isolated by gaps is a classic exhaustion-island turn — high conviction when it appears at a trend extreme.
2. Rare. Full two-sided gaps are uncommon in 24/7 markets (crypto), more frequent on gapping session markets (equities).
3. Confirm with location. Strongest at a tested high (bearish) or low (bullish).

Common pitfalls

• Body-only gaps. This pattern needs full high/low gaps, not just body gaps — stricter than morning/evening star.
• 24/7 markets. Perpetual crypto rarely gaps; expect few prints there.

References

• Steve Nison, Japanese Candlestick Charting Techniques (1991).

See also

• MorningDojiStar / EveningDojiStar — same shape with body gaps only.
• DojiStar — the two-bar doji-gap precursor.
• Indicators-Overview — the full taxonomy.