Rocr

Rate of Change Ratio (ROCR) — close / close[period], the momentum ratio relative to the price period bars ago.

Quick reference

Field	Value
Family	Momentum Oscillators
Input type	f64 (single close)
Output type	f64
Output range	> 0; 1 = no change, > 1 advance, < 1 decline
Default parameters	period is required
Warmup period	period + 1
Interpretation	A multiplicative momentum ratio: 1.05 means price is 1.05× its value period bars ago.

Formula

For t >= period (after warmup):

ROCR = close_t / close_{t-period}

This is Rocp plus one (ROCR = 1 + ROCP): the ratio form centred on 1 rather than 0. It is convenient for compounding, since chaining two ratios multiplies them. Where the reference price close_{t-period} is zero the result is guarded to 0. See crates/wickra-core/src/indicators/rocr.rs.

Parameters

Name	Type	Default	Valid range	Description	Source
period	usize	none	>= 1	Lookback distance in bars. period = 0 errors with Error::PeriodZero.	rocr.rs:40

Inputs / Outputs

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

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

Python streams as float | None, batches as a 1-D numpy.ndarray (NaN for warmup). Node streams as number | null, batches as Array<number> with NaN placeholders.

Warmup

Rocr::new(period).warmup_period() == period + 1 (the accessors_report_config test pins warmup_period() == 4 for period = 3). The first value is emitted at input index period — the (period + 1)-th input — once a reference price period bars back is available. The known_value_is_a_ratio test pins out[0] == None.

Edge cases

• Zero reference price. When close[period] == 0 the division is guarded and the output is 0. The unit test zero_reference_price_reports_zero pins this.
• Constant series. A flat price returns 1 for every emitted value. The unit test constant_series_yields_one pins this.
• Non-finite input. NaN / infinity inputs are ignored: they leave the window untouched and the last computed value is returned. The unit test non_finite_input_holds_last pins this.
• Reset. r.reset() clears the window and last value. The unit test reset_clears_state pins this.

Examples

Rust

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

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let mut r = Rocr::new(1)?;
    let out: Vec<Option<f64>> = r.batch(&[10.0, 11.0]);
    println!("{:?}", out);
    Ok(())
}

Output:

[None, Some(1.1)]

Period 1 over [10, 11]: 11 / 10 = 1.1. This matches the known_value_is_a_ratio unit test in crates/wickra-core/src/indicators/rocr.rs.

Python

import numpy as np
import wickra as ta

r = ta.Rocr(1)
print(r.batch(np.array([10.0, 11.0])))

Output:

[nan 1.1]

Node

const ta = require('wickra');
const r = new ta.Rocr(1);
for (const x of [10, 11]) console.log(x, '->', r.update(x));

Output:

10 -> null
11 -> 1.1

Interpretation

Rocr expresses momentum as a multiplier rather than a difference. Read 1.0 as the flat line: above it price has advanced over the lookback, below it price has declined. Because it is a ratio it composes naturally — multiplying successive non-overlapping ROCR values reconstructs the cumulative price ratio. It is unbounded above and bounded below by 0, so it is asymmetric around its 1.0 pivot (a halving reads 0.5, a doubling reads 2.0).

Choose Rocr over Rocp when you want a ratio centred on 1; choose Rocr100 when you prefer the same ratio rescaled so the flat line sits at 100.

Common pitfalls

• Asymmetry around 1. Equal-magnitude up and down moves are not symmetric: +10% is 1.10 but −10% is 0.90, and a move that doubles then halves returns to 1.0 only because ratios multiply, not add. Do not average ROCR values arithmetically.
• Reference-price zero. A guarded 0 (not NaN) is returned when the lookback price is exactly zero.

References

The Rate of Change Ratio is TA-Lib's ROCR function; it is the multiplicative form of the Rate of Change momentum oscillator.

See also

• Indicator-Rocp — the fractional form (ROCR − 1).
• Indicator-Rocr100 — the same ratio ×100.
• Indicator-Roc — the percentage form.
• Indicators-Overview — the full taxonomy.