Skip to content

TruncatedBandpass

Ehlers DSP filter ehlers dsp bandpass cycle robust

Truncated Bandpass filter for handling sharp price movements.

Visual Example

TruncatedBandpass — annotated preview mapping to core implementation

Synthetic ideal per library logic. Generated 2026-06-25 IST via docs/generate_all_previews.py (reproducible; maps to core Next<T> implementation).

Description

The TruncatedBandpass indicator is a technical analysis tool that truncated bandpass filter for handling sharp price movements.

This indicator is primarily used for identifying key market conditions. It provides a robust signal that can be easily integrated into both simple strategies and more complex machine learning feature pipelines. Compared to its alternatives, it offers a distinct balance of responsiveness and stability.

Traders often combine this with other metrics to confirm signals and avoid false positives during sideways market regimes. It remains a standard tool for systematic trading models.

Use to isolate cyclic components while minimizing 'ringing' effects caused by sudden price shocks. Ideal for cycle-based trading systems in volatile markets.

Finite Impulse Response (FIR) filters have a fixed history, while Infinite Impulse Response (IIR) filters technically have an infinite history. Truncation limits the IIR feedback loop to a specific length, combining the sharp selectivity of IIR with the outlier-rejection of FIR.

QuantWave implements this indicator via the universal Next<T> trait, guaranteeing bit-identical results between Rust streaming, Python streaming, and Polars batch (.ta() / map_batches) surfaces.

Formula / Specification

Implementation (quantwave-core/src/indicators/truncated_bandpass.rs):

[ L1 = \cos(360/P), \quad G1 = \cos(BW \cdot 360/P), \quad S1 = 1/G1 - \sqrt{1/G1^2 - 1} ] [ BPT_t = \text{IIR window of length } L \text{ with zero initial conditions} ]

Gold-standard parity vectors: quantwave-core/tests/gold_standard/truncated_bandpass.json.

Parameters

Parameter Default Description
period 20 Cycle period to isolate
bandwidth 0.1 Bandwidth of the filter
length 10 Truncation length

Usage Examples

Streaming (Rust)

use quantwave_core::indicators::TRUNCATED_BANDPASS;
use quantwave_core::traits::Next;

let mut ind = TRUNCATED_BANDPASS::new(20);
for price in &prices {
    let value = ind.next(price);
}

Streaming (Python)

from quantwave import TRUNCATED_BANDPASS

ind = TRUNCATED_BANDPASS(20)
for price in prices:
    value = ind.next(price)

Polars Batch (Python)

import polars as pl
import quantwave as qw

def apply_truncatedbandpass(series: pl.Series) -> pl.Series:
    ind = qw.TRUNCATED_BANDPASS(20)
    return pl.Series([ind.next(float(v)) for v in series.to_list()])

df = (
    pl.read_csv('ohlcv.csv')
    .lazy()
    .with_columns(
        pl.col("close").map_batches(apply_truncatedbandpass, return_dtype=pl.Float64).alias("truncatedbandpass")
    )
    .collect()
)

All surfaces are bit-identical via the single Next<T> implementation and proptests.

Edge Cases & Limitations

  • Recursive DSP filters require a warm-up period; first N bars may be unstable or raw-pass-through.
  • Designed for cyclic/mean-reverting regimes; trending markets can produce lag or drift.
  • Parameter period (or equivalent) controls cutoff — too small adds noise, too large adds lag.
  • Prefer chaining with other Ehlers tools (Roofing Filter, SuperSmoother) on noisy inputs.
  • Validated via proptests against gold-standard vectors where available.
  • No look-ahead bias; suitable for live streaming and batch feature pipelines.

Boundary Behavior

Condition Behavior
Warm-up Leading bars return NaN until warmup_bars is satisfied.
period > len When period exceeds series length, output is all NaN.
NaN inputs NaN in input propagates to output (NaN out).
Invalid params Non-positive period or missing required params raise ValueError.
Empty data Empty input returns an empty result series.

Sources & References

Primary Source: https://www.traders.com/Documentation/FEEDbk_docs/2020/07/TradersTips.html

Implementation: quantwave-core/src/indicators/truncated_bandpass.rs (TRUNCATED_BANDPASS / TRUNCATED_BANDPASS_METADATA). Parity: quantwave-core/tests/gold_standard/truncated_bandpass.json

Provenance: Standards bulk upgrade 2026-06-25 IST — see docs/DOCUMENTATION_STANDARDS.md.