R help - Sep 2004 - strptime problems

Hi, I'm experiencing a problem with strptime. (R 1.9.1 on a Win2000
machine)
I have a large list containing 6 columns and 161800 rows.
One column contains dates that I want to convert in order to compare the
different dates.
Some dates work just fine while others become NA. I don't see any
difference between the dates.
I've attached an example from my code. Hope this explains my problem.
Thanks in advance
/Per
 
 
> ccc[[11]]$TradeDateTime[161521]
[1] 29-Sep-00:16:29:50
227345 Levels: 04-Jan-00:09:29:00 04-Jan-00:09:30:00 04-Jan-00:09:31:00
04-Jan-00:09:32:00 04-Jan-00:09:33:00 04-Jan-00:09:34:00 ...
30-Nov-99:23:55:00
>
strptime(as.character(ccc[[11]]$TradeDateTime[161521]),"%d-%b-%y:%H:%M:%
S")
[1] "2000-09-29 16:29:50"
#This works fine!
 
> ccc[[11]]$TradeDateTime[161522]
[1] 19-May-99:11:38:00
227345 Levels: 04-Jan-00:09:29:00 04-Jan-00:09:30:00 04-Jan-00:09:31:00
04-Jan-00:09:32:00 04-Jan-00:09:33:00 04-Jan-00:09:34:00 ...
30-Nov-99:23:55:00
>
strptime(as.character(ccc[[11]]$TradeDateTime[161522]),"%d-%b-%y:%H:%M:%
S")
[1] NA
#This doesn't

	[[alternative HTML version deleted]]

Per Wiklund wrote:

 >
 > Hi, I'm experiencing a problem with strptime. (R 1.9.1 on a Win2000
 > machine)

[...]

 > strptime(as.character(ccc[[11]]$TradeDateTime[161522]),
 >    "%d-%b-%y:%H:%M:%S")
 > [1] NA

[...]

Hello,

unfortunately, your code does not show what data are in the object ccc, 
so an exact diagnosis is almost impossible. However, I suspect you have 
timezone problems which are common on Windows computers. If you don't 
need timezone calculations I suggest to use the Date class (uppercase 
"D") or the chron package instead of the POSIX classes, see the last 
issue of R-News:

http://cran.r-project.org/doc/Rnews/Rnews_2004-1.pdf

Thomas P.

Thank you very much for your answer, the suggested locale change solved
the problem.
 
/Per

<-----Original Message----->

 	  	 From: Prof Brian Ripley
Sent: 9/3/2004 3:35:41 PM
To: petzoldt@rcs.urz.tu-dresden.de
Cc: par@wiklund.net;r-help@stat.math.ethz.ch
Subject: Re: [R] strptime problems 

I suspect rather a locale problem which Date and chron will also have. 

You are looking for an abbreviated month name (%b). What language is the
object ccc in? What language is your computer in? See the example in 
?strptime 

## read in date info in format 'ddmmmyyyy' 
## This will give NA(s) in some locales; setting the C locale 
## as in the commented lines will overcome this on most systems. 
## lct <- Sys.getlocale("LC_TIME");
Sys.setlocale("LC_TIME", "C")
x <- c("1jan1960", "2jan1960", "31mar1960",
"30jul1960")
z <- strptime(x, "%d%b%Y") 

Looking for English abbreviations in a Norwegian file, for example, will
not work and give NAs. 

On Fri, 3 Sep 2004, Thomas Petzoldt wrote: 
> Per Wiklund wrote: 
> 
> > 
> > Hi, I'm experiencing a problem with strptime. (R 1.9.1 on a
Win2000
> > machine) 
> 
> [...] 
> 
> > strptime(as.character(ccc[[11]]$TradeDateTime[161522]), 
> > "%d-%b-%y:%H:%M:%S") 
> > [1] NA 
> 
> [...] 
> 
> Hello, 
> 
> unfortunately, your code does not show what data are in the object
ccc, 
> so an exact diagnosis is almost impossible. However, I suspect you
have 
> timezone problems which are common on Windows computers. If you don't 
> need timezone calculations I suggest to use the Date class (uppercase 
> "D") or the chron package instead of the POSIX classes, see the
last
> issue of R-News: 
> 
> http://cran.r-project.org/doc/Rnews/Rnews_2004-1.pdf 
> 
> Thomas P. 
-- 
Brian D. Ripley, ripley@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 

. 
	

	[[alternative HTML version deleted]]