Shape · Time series

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.

Also known as: Sales forecast, time-series projection

How it works

Choose Output: Forecast predicts the periods ahead, Backtest holds out the most recent periods and forecasts them from the rest. Pick optional Group by columns (one forecast per group), the Date column, the Value column, a Frequency (Auto-detect reads it from the gaps between dates) and the Horizon in periods, 6 by default; a backtest's Hold out follows the horizon unless set. Method Auto compares exponential smoothing against seasonal naive on a hold-out and keeps smoothing only when it wins. Interval is 80% or 95% (default). The input must be one row per period per group with no gaps, duplicates or NULLs; otherwise the node stops with a message naming the fix, usually Resample first. The output keeps the group, date and value columns and adds forecast, lower, upper and is_forecast; untick Include history rows to keep only the future rows. A backtest adds naive_forecast, method, error and abs_pct_error.

Inputs and outputs

Settings

{
  dateColumn: string
  valueColumn: string
  groupBy?: string[]
  frequency?: "auto"|"day"|"week"|"month"|"quarter"|"year"
  horizon: number
  method?: "auto"|"seasonal-naive"|"exponential-smoothing"
  interval?: 80|95
  includeHistory?: boolean
  output?: "forecast"|"backtest"
  holdOut?: number
}

Example

Next quarter of monthly sales by region: date column month, value column sales, group by region, frequency Month, horizon 3. To see how well that would have worked last quarter, switch Output to Backtest with hold out 3.

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