Shape · Time series

Exponential Weighted Mean

Smooths a number column so recent rows count more than older ones. Gives a weighted mean, standard deviation or variance.

Also known as: EWMA, exponential smoothing

How it works

Group by computes one series per group. Order by sorts the rows first; with no sort the input order is used. Pick the Column to smooth and how to Decay by: Span, Alpha or Halflife, with one value; exactly one of the three must be set. Choose the Operation, Mean, Std or Var. The weights follow the recursive form (adjust off), so each value is alpha * current + (1 - alpha) * previous; rows where the source is NULL stay NULL in the output. The result is appended as column_ewm_span (or _alpha, _halflife) unless you set an Output column. This node runs on the desktop with the whole table in memory.

Inputs and outputs

Settings

{
  column: string
  decayType: "span"|"alpha"|"halflife"
  span?: number
  alpha?: number
  halflife?: number
  operation: "mean"|"std"|"var"
  outputColumn?: string
  groupBy?: string[]
  orderBy?: [{ column, direction }]
}

Example

A 20-day exponential moving average of a closing price per ticker: column close, decay by Span with value 20, operation Mean, order by day, group by ticker, output ema_20.

More in Time series

• Shift / Lag — Copies a value from N rows earlier or later onto each row, so you can compare a row with the one before it. For example, last month's sales next to this month's.
• Diff / Pct Change — Compares each row with the one N periods earlier and writes the difference or percent change. For example, month-over-month growth per store.
• Rolling Window — Adds moving statistics over the last N rows: sum, average, standard deviation, min, max, count or median. For example, a 7-day moving average.
• Resample — Rolls a time series up into regular buckets, from minutes to years, with a total or other summary per bucket. Empty buckets can be left out, carried forward or set to zero.
• Forecast — Forecasts the next periods of a measure from a dated series, one forecast per group, with a lower and upper band. It can also backtest the recent past to show how accurate the forecast would have been.
• Sessionize — Groups events into sessions: a new session starts when the gap since the last event is longer than you set, such as 30 minutes.