[R] time zone - best way to shift hours

[Prof Brian Ripley]

On Thu, 24 Jul 2008, DCC wrote:

>
> I frequently work with hourly data in GMT format, and ensure that any data I
> read into R are expressed in GMT through as.POSIXct(mydate, tz = "GMT").

That makes no sense.  POSIXct times are absolute times, not in any 
timezone.  You can display them in any timezone you like, e.g. by the 
format method.

If you have a character representation of a time to convert by as.POSIXct, 
you do need to tell R the timezone used in the character representation.

> I am interested in processing air pollution data in GMT where where peaks in
> emissions (say from road traffic sources) tend to occur at the same LOCAL
> time each day.  I am interested in calculating mean concentrations by hour
> of day i.e. a diurnal profile, consisting of 24 hourly means.  The profile
> would be clearer if all hours were in local time e.g. the daylight saving
> hours were shifted by +1 hour.  This would ensure that the summer and winter
> profiles match up with those in emissions.
>
> I've looked though the various R documents/forum, but can't seem to identify
> the easiest approach to this.
>
> Any suggestions?

I think you want

> T <- Sys.time()
> as.POSIXlt(T, tz="Europe/London")$hour

which gives you the hour in the local timezone of Leeds.

>
> Many thanks.
>
> David

> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595