Thanks for confirming that it's a wider problem. I thought that /etc/localtime
on my machine was the Olsen database, but presumably not, since the DST changes
are correct in that file. Unless R is reading that file incorrectly?
I've ended up using chron (which ignores time zones and DST) and a
What OS are you using? It makes a difference.
---
Jeff NewmillerThe . . Go Live...
DCN:jdnew...@dcn.davis.ca.usBasics: ##.#. ##.#. Live Go...
On 06.12.2012 15:54, Jeff Newmiller wrote:
What OS are you using? It makes a difference.
... and R version.
Uwe Ligges
---
Jeff NewmillerThe . . Go Live...
Apologies - I intended to add that info at the end: OS X 10.8.2 and R 2.15.1.
Many thanks.
R.Version()
$platform
[1] x86_64-apple-darwin9.8.0
$arch
[1] x86_64
$os
[1] darwin9.8.0
$system
[1] x86_64, darwin9.8.0
$status
[1]
$major
[1] 2
$minor
[1] 15.1
$year
[1] 2012
$month
[1] 06
$day
[1] 22
I see this discontinuity at '2010-04-04 02:26:05' on my Windows 7 x64 with R
2.15.2. I also see it for time zone America_Los_Angeles at '2010-11-07
01:23:09'.
I am not entirely sure what the correct behavior should be, since the string
does not specify whether it is in daylight or standard
Can anyone please shed any light on why R DateTimeClasses give weird times for
when daylight saving time information changes, and which aren't consistent with
the OS?
Example:
Expected result: in New Zealand DST stopped (NZDT - NZST) at 03:00
NZDT on 2010-04-04, as confirmed by the
6 matches
Mail list logo