[R] Unexpected behavior of clocktime related to daylight savings time

[Thomas Lumley]

I'm not surprised that strange things happen if you try to handle times in 
the duplicated hour without specifying which time you mean.  If your 
current time zone setting is something like EST5EDT that includes both 
daylight-saving and standard times then  2008-11-02 01:16:00 occured 
twice, so some of the calls to difftime() are ambiguous.  If your 
time-zone setting specifies daylight-saving time or not, then some of the 
calls to difftime() involve times that did not exist.  If you create 
POSIXt objects that unambiguously specify the times then things should be 
more predictable.

 	-thomas

On Fri, 7 Nov 2008, Dennis Fisher wrote:

> Colleagues,
>
> I submitted this several days ago and no one responded, so I am trying
> again, trying a different subject line:
>
> I just encountered some unexpected behavior of difftime in
> relationship to the change from daylight savings to standard time.
>
> My understanding is that DST and ST take effect at 2AM.  However, the
> results below suggests that R (version 2.8.0 in OS X) implements the
> change at 2:16AM:
>
> Expected:
>>> difftime("2008-11-02 02:01:00", "2008-11-02 00:59:00")
>> Time difference of 2.033333 hours
>>> difftime("2008-11-02 01:16:00", "2008-11-02 01:15:00")
>> Time difference of 1 mins
>>> difftime("2008-11-02 01:18:00", "2008-11-02 01:17:00")
>> Time difference of 1 mins
>
> Not expected:
>>> difftime("2008-11-02 01:17:00", "2008-11-02 01:16:00")
>> Time difference of 1.016667 hours
>
> Can anyone explain this?
>
> Dennis
>
>
> Dennis Fisher MD
> P < (The "P Less Than" Company)
> Phone: 1-866-PLessThan (1-866-753-7784)
> Fax: 1-415-564-2220
> www.PLessThan.com
>
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle