Apply Ideas¶
Technique: Observe system behavior over time using rolling windows or moving averages.
Rolling calculations help smooth short-term variation and reveal longer-term patterns.
A good dataset for this module:
- contains observations ordered over time
- includes repeated measurements
Example Systems¶
Website Visits¶
Possible fields:
- timestamp
- visits
Questions to explore:
- What does a rolling average reveal about traffic trends?
- Are there periods of unusually high activity?
Energy Demand¶
Possible fields:
- timestamp
- demand_mw
Questions to explore:
- Do rolling averages reveal peak usage patterns?
- Are there seasonal or daily cycles?
Public Transit Ridership¶
Possible fields:
- date
- rides
Questions to explore:
- How does ridership change across days or weeks?
- Do rolling averages reveal longer trends?
Application Response Time¶
Possible fields:
- timestamp
- latency_ms
Questions to explore:
- Does system performance degrade during certain periods?
- Do rolling metrics reveal slow trends not visible in individual observations?
Temperature Monitoring¶
Possible fields:
- timestamp
- temperature
Questions to explore:
- How do rolling averages change across time?
- Are there patterns across days or seasons?