Time

In R, dates and times are special data types (classes), not just text strings. They contain information about order (which observation comes before another) and distance, and also the position in cycles of year, month, weeks. A characterstring may look like a date, but has a different class.

c1 <- "2024-01-01"
c2 <- "2024-01-02"
class(c1); class(c2)
#c2-c1 results in error

here, we cannot calculate the difference, but it will work with date/time formats.

R mainly uses two types of time-related classes:

Daterepresents a calender date - naturally, the resolution is days.

d1 <- as.Date("2024-01-01")
d2 <- as.Date("2024-01-02")
class(d1); class(d2)
d2-d1

Internally, dates are represented as the number of days since 1970-01-01, with negative values for earlier dates.

POSIXct represents date and time. Here, the resolution is seconds.

t <- as.POSIXct("2024-01-01 12:30:00", tz = "UTC")
class(t)

Date formats

A format describes how a date is written as text.

Format	Example
%Y-%m-%d	2024-01-31
%d.%m.%Y	31.01.2024
%m/%d/%Y	01/31/2024
%Y-%m-%d %H:%M:%S	2024-01-31 14:00:00

Converting

Convert to date or time using as.Date() or as.POSIXct(), respectively. If your character is not in the standard form, specify the format:

as.Date("31.01.2024", format = "%d.%m.%Y")

Attention: as.POSIXct assumes the timestamp is given in UTC. You can overwrite that by adding the argument tz

t <- as.POSIXct("2024-01-01 12:00:00", tz = "UTC")
t
t_berlin <- as.POSIXct("2024-01-01 12:00:00", tz = "Europe/Berlin")
t_berlin

To convert between timezonesand given a defined timezone, you can again use as.POSIXct:

t_berlin_utc <- as.POSIXct(t_berlin, tz="UTC")
t_berlin_utc

Extracting components

You can extract calender components using as.POSTIXlt()

tt <- as.POSIXlt()

This creates a list-like object with named components.

Component	Code	Meaning
Hour	tt$hour	0–23
Day of week	tt$wday	0–6 (Sunday = 0)
Month	tt$mon	0–11
Day of month	tt$mday	1–31
Year	tt$year	years since 1900