Gnumeric · Forecasting · Intermediate

Smooth Measurements with a Moving Average in Gnumeric

Compare trailing and centered moving averages in Gnumeric with a simple series, expected values and common chart mistakes.

Official documentation reviewed; native application verification pending · Updated 10 October 2026 · Editorial standards

A moving average reduces short-term variation by averaging neighbouring observations. Gnumeric’s Moving Average analysis tool supports simple, cumulative, weighted and Spencer’s 15-point methods in the official manual.

Practice series

Download weekly readings (CSV). The eight consecutive values are 12, 14, 16, 18, 20, 22, 24 and 26.

For a trailing three-period simple moving average, the first complete average is:

(12 + 14 + 16) / 3 = 14

The next is (14 + 16 + 18)/3 = 16, then 18, 20, 22 and 24. It is normal for the first two rows of the moving-average output to have no complete three-period window.

Choose trailing or centered windows

  1. Open the data and select the numeric column without including unrelated headers as observations.
  2. Find Moving Average in Gnumeric’s statistical-analysis tools.
  3. Choose Simple moving average and an Interval of 3.
  4. Select a prior/trailing window when you want a value based on the current and previous two readings.
  5. Output to a new sheet and, if useful, include a chart comparing raw data and the smoothed curve.

A centered three-point average at a time position uses the previous, current and next observation; it changes the alignment and uses future observations. Do not treat a centered smoother as a real-time forecast.

Which average should you use?

  • Simple/trailing: useful for a transparent rolling indicator.
  • Cumulative: averages all readings from the beginning to the current observation.
  • Weighted: gives different weights to observations; understand how your selected method sets those weights.
  • Spencer’s 15-point: needs much longer series; do not choose it for our eight-row example.

Moving averages do not establish statistical significance and may hide abrupt outliers. Always check the raw measurements.

Verification note: The example arithmetic is independently verified. Gnumeric-specific menu behavior needs native checking.


Official source: Application handbook or manual. This guide is an editorial draft prepared for staging; exact menus and outputs must be checked in the target software/version before public release.