Time series attributes

Manipulating Time Series Data in R

Harrison Brown

Graduate Researcher in Geography

Temporal attributes

Start point
End point
Frequency

Temporal attributes can be used to:

Understand when and for how long the data was sampled
Find the times where multiple time series overlap

Conceptual diagram of two overlapping time series; the last part of time series A overlaps the first part of time series B.

Conceptual diagram outlining the challenge of combining two overlapping time series. In the overlapping region of the combined time series, there is a dotted line, representing how the data in that region is uncertain.

Temporal attributes

Key temporal attributes:

Start point

start(AirPassengers)

[1] 1949 1

Temporal attributes

Key temporal attributes:

Start point
End point

end(ftse)

[1] 1998.646

Decimal date (numeric)

end(ftse) %>%
  lubridate::date_decimal()

[1] "1998-08-24 20:18:27 UTC"

Temporal attributes

Key temporal attributes:

Starting point
Endpoint
Frequency

frequency(monthly_data)

[1] 12

frequency(weekly_data)

[1] 52.14286

frequency(ftse)

[1] 260

ftse

1991.4962 2443.6
1991.5    2460.2
1991.5038 2448.2
1991.5077 2470.4
1991.5115 2484.7
1991.5154 2466.8
...

Regular vs. irregular time series

A conceptual timeline showing a regular time series. Above the line are markers representing each observation. The markers are spaced at regular intervals.

Evenly-spaced intervals
No missing values
Uses decimal date for 'irregular' intervals

A conceptual timeline showing an irregular time series. As opposed to the 'regular time series' diagram, the markers in this time series are not spaced at regular intervals, with uneven gaps between observations.

Spacing can be irregular
- Weekdays, random days
Missing observations
Decimal date or Date/POSIXct data

Let's practice!

Manipulating Time Series Data in R