ProfileShape

Classifies the rolling volume profile by where its point of control sits: P-shape (+1), b-shape (−1), or D/normal (0).

Quick reference

Field	Value
Family	Market Profile
Input type	Candle (high / low / volume)
Output type	f64 (shape code)
Output range	{−1, 0, +1}
Default parameters	(period = 20, bins = 24) (Python)
Warmup period	period
Interpretation	+1 P-shape (short-cover/accum); −1 b-shape (liquidation); 0 balanced.

Formula

build a `bins`-bucket volume profile over the last `period` candles
poc_idx = bin with the most volume
+1  P-shape : POC in the upper third
−1  b-shape : POC in the lower third
 0  D/normal: POC in the middle third

Market Profile readers classify the day by where the heaviest trading sits. A P-shape (control high, thin tail below) marks short-covering or early accumulation; a b-shape (control low, thin tail above) marks long liquidation or distribution; a D-shape is a balanced two-sided day. Source: crates/wickra-core/src/indicators/profile_shape.rs.

Parameters

Name	Type	Default	Valid range	Source	Description
period	usize	20 (Python)	>= 1	profile_shape.rs:60	Rolling profile window. 0 errors with Error::PeriodZero.
bins	usize	24 (Python)	>= 3	profile_shape.rs:60	Price buckets (need three for the split). < 3 errors with Error::InvalidPeriod.

params() returns (period, bins); value returns the current code if ready.

Inputs / Outputs

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

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

A Candle in, an Option<f64> out (+1/0/−1). Python update(candle) / batch(high, low, volume); Node update(high, low, volume) / batch(...).

Warmup

warmup_period() == period. The profile window must fill first (first_emission_at_warmup_period pins this).

Edge cases

• P-shape. Heavy top volume → +1 (heavy_top_is_p_shape pins this).
• b-shape. Heavy bottom volume → −1 (heavy_bottom_is_b_shape pins this).
• D-shape. Balanced middle → 0 (balanced_is_d_shape pins this).
• Finiteness. Candle::new rejects non-finite fields, so no in-method guard is needed.
• Reset. p.reset() clears the window and the last value (reset_clears_state).

Examples

Rust

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

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut p = ProfileShape::new(6, 9)?;
    let c = |hi: f64, lo: f64, v: f64| Candle::new((hi+lo)/2.0, hi, lo, (hi+lo)/2.0, v, 0).unwrap();
    let mut candles: Vec<Candle> = (0..5).map(|_| c(119.0, 117.0, 5_000.0)).collect();
    candles.push(c(119.0, 80.0, 50.0)); // thin tail down
    println!("{:?}", p.batch(&candles).last().unwrap()); // Some(1.0) P-shape
    Ok(())
}

Output:

Some(1.0)

Python

import numpy as np
import wickra as ta

p = ta.ProfileShape(20, 24)
high = ...; low = ...; volume = ...
print(p.batch(high, low, volume)[-1])  # -1 / 0 / 1

Node

const ta = require('wickra');
const p = new ta.ProfileShape(20, 24);
console.log('warmupPeriod:', p.warmupPeriod()); // 20

Streaming

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

let mut p = ProfileShape::new(20, 24).unwrap();
let feed: Vec<Candle> = Vec::new(); // your live stream
for candle in feed {
    match p.update(candle) {
        Some(1.0)  => println!("P-shape (short-covering / accumulation)"),
        Some(-1.0) => println!("b-shape (long liquidation / distribution)"),
        Some(0.0)  => println!("D-shape (balanced)"),
        _ => {}
    }
}

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

Interpretation

1. Posture. P after a downtrend = short-covering (possible bottom); b after an uptrend = distribution (possible top).
2. Continuation vs. reversal. Read the shape in context of the prior trend to judge whether it confirms or warns.
3. Balance detection. Repeated D-shapes signal a balancing market — fade the edges until it breaks.

Common pitfalls

• Coarse code. Three buckets summarise a rich profile; inspect the full VolumeProfile for nuance.
• Window = session. Set period to one session's bars for classic shape reads.
• Bins resolution. Too few bins coarsen the POC location.

References

Dalton, J. F. (1993), Mind Over Markets; Steidlmayer, J. P. — Market Profile.

See also

• Indicator-VolumeProfile — the full profile.
• Indicator-HighLowVolumeNodes — HVN/LVN levels.
• Indicator-ValueArea — the 70% value area.
• Indicators-Overview — the full taxonomy.