When working with multiple time series, joining them typically involves aligning records by timestamp. This can serve two purposes:

  • Combine metrics: Add additional measurements from different sources.
  • Extend time span: Append data to create a longer series.

Key Consideration: Granularity

  • The time granularity of each series matters when joining.
  • If one series has hourly data and another has daily data, you must decide how to aggregate or resample before the join.

Clock Skew

  • Clock skew occurs when timestamps do not perfectly align across sources.
  • A common solution is to truncate or round timestamps to the desired precision (e.g., nearest minute, hour).

Related Analysis Patterns