|
Category
Multi-Bar Functions
Description
Ultimate Smoother filter as described by John Ehlers.
Syntax
EhlersUS(expr, count)
Parameters
expr - data series formula
count - lookback period
Notes
Combines HighPass and SuperSmoother characteristics to remove high-frequency noise while preserving the underlying trend with very little lag.
Produces a smoothed version of the source series on the same scale as the input, preserving the overall trend.
The count argument may be a formula rather than a constant. A non-constant period is re-evaluated on every bar: the filter coefficients are recomputed from that bar’s value and the recursion continues from the stored prior outputs, the same adaptive model used by AEMA. The period formula may even reference the enclosing Data item’s own prior values at offsets of 1 or more, allowing self-tuning filters.
On bars where the period evaluates to n/a, the output is n/a, and the recursion re-seeds when the period next becomes valid. The effective period is internally clamped to a minimum of 2 bars.
This function supports one-pass calculation when used in the Data Section.
Example
Smooths the closing price over roughly a 20-bar period, removing short-term noise with very little lag while preserving the overall trend.
The result stays on the same scale as price and can be used as a low-lag trend or signal line.
|
