ETF5922 Chap.10 Time Series Visualisation
Time Series Visualisation
Time order is part of the data
Time series observations must be ordered and interpreted at their actual interval. A line implies continuity, so missing periods should remain visible rather than being bridged silently.
Gaps and intervals alter the story
Trend, seasonality, cycles, events and level shifts are distinct temporal structures.
Aggregation and moving summaries can reduce noise while concealing short-lived peaks.
Test a temporal claim
Connecting across an absent month visually invents continuity and can hide collection failure. A visible break preserves the distinction between an observed change and a period for which no value exists.
Test a temporal claim in context
Time series encode order and interval as part of meaning.
Aggregating daily observations to months changes variation and can smooth short events; indexing to a common start replaces magnitude with relative growth; connecting across a missing period visually invents continuity. Event annotations can locate coincidence but do not demonstrate impact without a suitable comparison.
A sound temporal explanation therefore identifies frequency, window, gaps, aggregation and baseline before naming trend or seasonality. Short sequences may support endpoint or month-to-month changes, but repeated seasonal claims require enough comparable cycles to distinguish recurrence from ordinary fluctuation.
What this chapter covers
- 01
Time series observations must be ordered and interpreted at their actual interval.
- 02
A line implies continuity, so missing periods should remain visible rather than being bridged silently.
- 03
Trend, seasonality, cycles, events and level shifts are distinct temporal structures.
- 04
Aggregation and moving summaries can reduce noise while concealing short-lived peaks.
- 05
Use timestamp and interval to frame the reader's task
- 06
Check chronology against trend before styling
- 07
Explain how seasonality changes the visible comparison
- 08
Audit cycle without removing necessary context
Calculate change without inventing continuity
- 2Calculate each month-to-month change in chronological order.
- 2Compute the first-to-last percentage using the first month as denominator.
- 2Keep a common vertical scale and label the observation interval.
- 2State that four months cannot establish annual seasonality.
Key terms
- Time index
- The ordered variable locating observations in time and defining which value precedes another.
- Seasonality
- A pattern that recurs at a meaningful and specified calendar period.
- Temporal gap
- An interval with no observation, displayed as missing rather than joined by an invented segment.
- Aggregation interval
- The span over which observations are combined, such as daily values converted into monthly totals.
Time Series Visualisation FAQ
Why should a missing month remain a gap?
Connecting across an absent month visually invents continuity and can hide collection failure. A visible break preserves the distinction between an observed change and a period for which no value exists.
When should a series be indexed?
Index values when relative change from a common starting point is the intended comparison and raw magnitudes would obscure that task. Label the base and retain access to original units. Indexing changes the question, so it should not replace a raw-level view when absolute scale matters.
How are trend and seasonality distinguished?
Trend describes longer-run directional movement, while seasonality requires a repeating pattern at a specified period. A short sequence may show rises and falls without covering enough cycles to establish seasonality. State the observation interval and window before applying either label.
What does a gap communicate in a line chart?
A gap states that no observation is available for that interval. Joining the neighbouring points instead suggests a measured path through the missing period. If interpolation is analytically justified, distinguish estimated values from observed ones and explain the method; otherwise preserve the break and report the completeness limitation.
Exam move
Plot a short monthly sequence containing one absent period and an event date. Compare raw levels, indexed change and a coarser aggregation while preserving the missing interval. For each version, state the denominator and temporal question. Finish by explaining why four months cannot establish annual seasonality even when the line appears to repeat or turn regularly.
Working through Time Series Visualisation in ETF5922? Sia is AskSia’s AI Computer Science tutor — ask any ETF5922 Time Series Visualisation question and get a clear, step-by-step explanation grounded in how ETF5922 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
