BollingerBandwidth

Summary

BollingerBandwidth is RTTA's streaming implementation of: Bollinger band width as (upper-lower)/middle for a rolling mean and standard-deviation envelope.

Update API

result = rtta.BollingerBandwidth().update(value)

The update(...) call consumes one observation using value. advance(...) uses the same inputs when the caller wants to update state without materializing a Python return value.

Theory Of Operation

BollingerBandwidth normalizes the distance between the upper and lower Bollinger bands by the middle band. Rising bandwidth marks volatility expansion; falling bandwidth marks compression.

Recurrence

Let \(z_t = value_t\), \(n\) the window, and \(k\) the standard-deviation multiplier (num_std, default 2).

\[M_t = \operatorname{mean}_n(z_t), \qquad S_t = \operatorname{stddev}_n(z_t)\]

\[BW_t = \frac{(M_t + k S_t) - (M_t - k S_t)}{M_t} = \frac{2 k S_t}{M_t}\]

The return value is the current scalar indicator value.

Composed Primitives

BollingerBands, SMA, StdDev

Implementation Notes

The recurrence is implemented in src/rtta/indicator.cpp in class BollingerBandwidth.

Reference

Background reference