← All workshopsRead Separating Repetition From Change

Time Series and Forecasting

Measure change across matching cycle positions

Remove an exactly repeating cycle

Botie wants to see demand changes after comparing matching positions in successive cycles.

BOTIE

The demand sequence is [5, 9, 5, 9] with period 2. Which seasonal differences remain?
The inputs we are comparing

Chronological demand

Two-position cycle

  1. 010
  2. 120
  3. 212
  4. 322

Exact repetition

  1. 05
  2. 19
  3. 25
  4. 39

Seasonal period in observations

Same in both cases

2
Result for two-position cycle
  1. 02
  2. 12

For a list use brackets, such as [2, 4]. For a matrix put each row in brackets. Text answers can stay as text.

A three-day pattern changes unevenly

This check uses the pieces from “Subtract aligned seasonal references”.

BOTIE

Botie sees a repeating three-day delivery cycle. Compare every observation in the second cycle with the matching position in the first cycle. Return the seasonal differences in time order.
Chronological demand
  1. 08
  2. 115
  3. 25
  4. 310
  5. 413
  6. 59
Seasonal period in observations
3

Use a number, text, or an array matching the requested result. For a matrix, put each row in brackets. Saving your prediction runs your current construction on these inputs and checks the earlier examples too.

Build a construction in the workspace above first. This case will test those pieces.

What would a zero seasonal difference tell you, and what would it leave unknown about future demand?

A supplied period defines the comparison; differencing neither discovers that period nor proves stationarity.

The period is supplied rather than discovered. Differencing alone does not establish stationarity.

Work through the Python challenges →
Continue the courseUpdating a Forecast Without Starting Over →
← Building Inputs From Earlier Days