R: Numeric Representation of Date Time

I’ve been playing around with date times in R recently and I wanted to derive a numeric representation for a given value to make it easier to see the correlation between time and another variable.

e.g. December 13th 2014 17:30 should return 17.5 since it’s 17.5 hours since midnight.

Using the standard R libraries we would write the following code:

> december13 = as.POSIXlt("2014-12-13 17:30:00")
> as.numeric(december13 - trunc(december13, "day"), units="hours")
[1] 17.5

That works pretty well but Antonios recently introduced me to the lubridate so I thought I’d give that a try as well.

The first nice thing about lubridate is that we can use the date we created earlier and call the floor_date function rather than truncate:

> (december13 - floor_date(december13, "day"))
Time difference of 17.5 hours

That gives us back a difftime…

> class((december13 - floor_date(december13, "day")))
[1] "difftime"

…which we can divide by different units to get the granularity we want:

> diff = (december13 - floor_date(december13, "day"))
> diff / dhours(1)
[1] 17.5
 
> diff / ddays(1)
[1] 0.7291667
 
> diff / dminutes(1)
[1] 1050

Pretty neat!

lubridate also has some nice functions for creating dates/date times. e.g.

> ymd_hms("2014-12-13 17:00:00")
[1] "2014-12-13 17:00:00 UTC"
 
> ymd_hm("2014-12-13 17:00")
[1] "2014-12-13 17:00:00 UTC"
 
> ymd_h("2014-12-13 17")
[1] "2014-12-13 17:00:00 UTC"
 
> ymd("2014-12-13")
[1] "2014-12-13 UTC"

And if you want a different time zone that’s pretty easy too:

> with_tz(ymd("2014-12-13"), "GMT")
[1] "2014-12-13 GMT"