The R Book

Dates and Times in R

The measurement of time is highly idiosyncratic. Successive years start on different days of the week. There are months with different numbers of days. Leap years have an extra day in February. Americans and Britons put the day and the month in different places: 3/4/2006 is March 4 for the former and April 3 for the latter. Occasional years have an additional ‘leap second’ added to them because friction from the tides is slowing down the rotation of the earth from when the standard time was set on the basis of the tropical year in 1900. The cumulative effect of having set the atomic clock too slow accounts for the continual need to insert leap seconds (32 of them since 1958). There is currently a debate about abandoning leap seconds and introducing a ‘leap minute’ every century or so instead. Calculations involving times are complicated by the operation of time zones and daylight saving schemes in different countries. All these things mean that working with dates and times is excruciatingly complicated. Fortunately, R has a robust system for dealing with this complexity. To see how R handles dates and times, have a look at Sys.time():

Sys.time()

[1] "2005-10-23 10:17:42 GMT Daylight Time"

The answer is strictly hierarchical from left to right: the longest time scale (years) comes first, then month then day separated by hyphens (minus signs), then there is a blank space and the time, hours first (in the 24-hour clock) then minutes, then seconds separated by ...